كل نماذج الذكاء الاصطناعي، في مكان واحد.
ابحث وقارن بين الأسعار والسرعة والخصوصية والسياق والقدرات قبل أن تختار النموذج المناسب.
375 نموذج مطابق
Fugu Max orchestrates our largest pool of models to push the cost–performance Pareto frontier. It dynamically finds efficient combinations of expert agents, aiming to improve model performance and cost efficiency together.
Fugu Ultra coordinates a deeper pool of expert agents to maximize answer quality on hard, high-stakes problems. Fugu Ultra focuses on maximizing performance, for a higher cost.
DeepSeek V4.1 Flash is an AI model from DeepSeek designed for fast, efficient multimodal interactions. It combines native multimodal capabilities with a new architecture focused on stronger performance, faster responses, and lower computational cost.
GPT Image 2.5 Flare generates images from text and image inputs, optimized for fast, high-quality everyday image generation.
GPT Image 2.5 Sunburst generates and edits images from text and image inputs, with a focus on precise editing.
Ling 3.0 Flash VL builds on Ling 3.0 Flash with stronger language capabilities, native visual perception, and visual agent capabilities. It supports text, image, and video inputs with text output, reasoning, and function calling.
Ling 3.0 Flash VL builds on Ling 3.0 Flash with stronger language capabilities, native visual perception, and visual agent capabilities. It supports text, image, and video inputs with text output, reasoning, and function calling.
Mercury 2.5 is Inception’s diffusion-based reasoning model for chat, agents, and structured workflows, with tool calling, structured outputs, and a 260K context window.
GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.
GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.
Ling-3.0-Flash-Sante is inclusionAI’s language model specialized for health and medicine, built on a Mixture-of-Experts architecture with 124 billion total parameters and approximately 5.1 billion active per token. With a 256K context window and function calling, it supports medical knowledge reasoning, evidence-based retrieval, and complex medical workflows while retaining general reasoning, coding, and agentic capabilities.
Ling-3.0-Flash-Sante is inclusionAI’s language model specialized for health and medicine, built on a Mixture-of-Experts architecture with 124 billion total parameters and approximately 5.1 billion active per token. With a 256K context window and function calling, it supports medical knowledge reasoning, evidence-based retrieval, and complex medical workflows while retaining general reasoning, coding, and agentic capabilities.
Gemini 3.8 Flash is the next iteration in the Gemini 3 model family, featuring algorithmic improvements to its core reasoning foundation. It supports customizable thinking configurations to control the mix of quality, cost and latency.
Speed-optimized version of Z.AI’s GLM-5.3 agentic coding model, built for responsive, real-time workloads. It supports advanced coding, long-horizon task execution, testing, vulnerability discovery, and cybersecurity analysis, making it ideal for developer tools, coding agents, and complex multi-step engineering workflows.
Qwen3.8-Max-0902 is an upgraded snapshot of qwen3.8-max.
Claude Fable 5.1 improves on Fable 5 across long-running agentic coding, knowledge work, and research, following instructions precisely over sessions that run unattended for hours.
Tencent Hy4 preview is Tencent Hy’s open-source large language model, featuring 770B total parameters, 49B active parameters, and a context window exceeding 1 million tokens. Built for real-world productivity, it supports long-horizon coding, cross-document analysis, office content creation, game development, and scientific reasoning, making it well-suited to complex agents and multi-step professional workflows.
Wan 3.0 Video Prime is Alibaba’s high-speed, all-in-one AI video generation model for creating polished clips from text, images, video, and audio references. It produces videos up to 30 seconds long with synchronized dialogue, music, and sound effects, while accelerated generation makes it ideal for rapid creative iteration and production workflows.
Ling 3.0 Flash Fin is InclusionAI’s finance-enhanced MoE language model, combining 124 billion total parameters with approximately 5.1 billion active parameters for efficient financial reasoning. Its 256K context window, function calling, and support for complex, multi-step investment workflows make it ideal for financial research, analysis, long-horizon planning, and execution, while retaining strong capabilities in coding and mathematics.
Ling 3.0 Flash Fin is InclusionAI’s finance-enhanced MoE language model, combining 124 billion total parameters with approximately 5.1 billion active parameters for efficient financial reasoning. Its 256K context window, function calling, and support for complex, multi-step investment workflows make it ideal for financial research, analysis, long-horizon planning, and execution, while retaining strong capabilities in coding and mathematics.
MiniMax H3 Max is a next-generation general-purpose multimodal video model ranking #1 for overall quality, prompt understanding, and aesthetics in all evaluations, while generating a 5-second video in under 3 seconds.
Transcription by google
Live transcription by Google
GLM-5.3-Flash is Z.ai’s native multimodal coding model, featuring 320B total parameters, 18B activated parameters, and a 1M-token context window. Its efficient hybrid attention architecture supports visual coding, tool use, and end-to-end professional workflows across code, browsers, documents, and graphical interfaces.
Muse Image is the first image generation model from Meta Superintelligence Labs, it uses advanced reasoning to understand complex prompts, seamlessly blending multiple photos into high-quality creations you can download and share anywhere.
Qwen3.8-Flash is Qwen’s fast, cost-efficient multimodal model, combining advanced reasoning and generation with a native 1M-token context window. Built for coding, agentic workflows, and visual understanding, it handles large codebases, long documents, charts, videos, and desktop applications.
All-in-one video generation model supporting text-to-video, image-to-video, first/last-frame, and omni-modal reference-based generation (image, video, and audio references) with synchronized audio, at up to 30 seconds per clip.
DeepSeek-V4-Flash-Vision-Exp is an experimental multimodal model that combines the agentic reasoning, coding, and world knowledge capabilities of DeepSeek-V4-Flash with advanced visual understanding. It delivers major gains on vision-dependent agent workflows while performing strongly across leading coding, repository, data science, automation, and multimodal benchmarks.
GLM 5.3 delivers comprehensive advancements in complex software engineering and agent capabilities. It uses the same base model as GLM-5.2, with all improvements driven by post-training.
Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability.
Gemini 3.7 Flash is the next iteration in the Gemini 3 model family, featuring algorithmic improvements to its core reasoning foundation. It supports customizable thinking configurations to control the mix of quality, cost and latency.
This is the 8/13 updated weights version of DeepSeek V4 Pro.
Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work. It stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application or work artifact.
NVIDIA Nemotron 3.5 Lightning is NVIDIA's fastest open model for always-on agents and high-volume specialized tasks. It delivers a substantial leap in agentic capability over its predecessor Nemotron 3 Nano, with up to 4x higher throughput on a 1M-token context.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Seedance 2.5 is a next-generation audio-video joint generation model, built for 30-second storytelling with precise reference control and powerful editing capabilities.
Ling 3.0 Flash is designed with token efficiency and production-scale agentic inference as key priorities, enabling developers to complete more useful work within constrained token, latency, and serving-cost budgets.
A coding-optimized model purpose-built for agentic workflows. Improvements to code generation, debugging, and codebase understanding — with a 1M context window that handles your entire project in one session.
A coding-optimized model with pricing designed for builders. Same model, same capabilities — up to 95% less than Standard. Your inputs and outputs are used to train and improve Meta's AI models.
FLUX 3 is a new multimodal foundation model that learns from images, videos, and audio within a unified architecture.
Open-weights release of the Qwen3.8 flagship (2.4T MoE, ~95B active); thinking always on with reasoning_effort low/medium/xhigh. The hosted Qwen 3.8 Max (vision, non-thinking, 1M default context) is Alibaba-only.
Sakana Namazu is a Japanese-specialized LLM that combines a deep understanding of Japanese culture and business customs with high-performance language capabilities. Built on the open model Kimi K2.6 and refined with Sakana AI's in-house data for Japanese language and business workflows, it handles complex tasks using web search and code execution. Unlike Fugu, which orchestrates multiple frontier models, Sakana Namazu provides a single in-house model as an API.
Qwen 3.8 Max is a 2.4-trillion-parameter MoE model delivering a comprehensive leap in coding and professional work. Autonomously codes and delivers complete projects spanning 10+ days. Handles hundreds of specialized tasks across legal, financial, design, and other professional domains, producing production-grade results end-to-end in a single conversation. Native visual understanding runs through the full cycle of planning, execution, and verification, enabling deep semantic analysis of ultra-long documents and extended video content. In long-horizon tasks, plans autonomously, iterates through closed feedback loops, and continuously evolves.
Inkling-Small is a lighter-weight model with 12B active parameters, trained with a similar recipe, to Inkling that achieves strong performance with even lower cost and latency.
H3 is a next-generation open-weights, general-purpose multimodal video model. Rather than being limited to specialized tasks such as generating, editing, or referencing, H3 understands multimodal contexts that bring together text, images, video, and audio. This enables it to interpret creative intent in a unified way and deliver more natural, coherent generation and expression.
Grok Voice Think Fast 2.0, our next-generation voice model with improved intelligence, transcription accuracy, and conversational capabilities.
The Qwen3.7 native vision-language Flash model series delivers a comprehensive upgrade over 3.6-Flash in multimodal understanding and agent execution. This model particularly excels in enhanced multimodal foundations with stronger universal object recognition, further improved real-world perception and spatial intelligence, significantly upgraded multimodal agent capabilities for Search Agent and CI Agent scenarios with more stable end-to-end task execution, as well as optimized multimodal coding for a smoother vibe coding experience.
S2.1 Pro is Fish Audio's current state-of-the-art voice model — the best model we have, now available to every developer for free via API. It is a neural speech synthesis model designed for production-grade AI voice generation, with particular strengths in low-latency streaming, multilingual TTS, and voice cloning.
S2.1 Pro is Fish Audio's current state-of-the-art voice model — the best model we have, now available to every developer for free via API. It is a neural speech synthesis model designed for production-grade AI voice generation, with particular strengths in low-latency streaming, multilingual TTS, and voice cloning.
Fast version of Kimi’s flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window.
Claude Opus 5 is the latest model in Anthropic's Opus family and a step-change improvement over Opus 4.8. It delivers major gains over Opus 4.8 in agentic coding, professional knowledge work, and long-horizon reasoning, and it is stronger per token across effort levels.
Claude Opus 5 is the latest model in Anthropic's Opus family and a step-change improvement over Opus 4.8. It delivers major gains over Opus 4.8 in agentic coding, professional knowledge work, and long-horizon reasoning, and it is stronger per token across effort levels.
Gemini 3.5 Flash Lite features upgraded agentic capabilities, making the model ideal for subagents in complex workflows.
Gemini 3.6 Flash delivers higher quality across coding, agentic workflows, and web development with reduced token consumption and fewer model calls compared to previous model iterations.
Laguna S 2.1 is Poolside's new open-weight model for agentic coding and long-horizon work.
Laguna S 2.1 is Poolside's new open-weight model for agentic coding and long-horizon work.
Kimi’s flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window.
Inkling is a multimodal MoE model (975B total, 41B active, 256k context) reasoning over text, image, and audio inputs.
Seedream-5.0-Pro, ByteDance's newest image generation model, delivers comprehensive upgrades for complex, lifelike image creation and editing, ushering in a new phase of controllable visual production. It stands out with precise editing control, robust commercial applicability and natural rendering results.
Fast response version of Kat Coder V2.5 optimized for Agent and Claw use cases.
KAT-Coder-V2.5 is a coding-focused agentic model trained to act autonomously inside real, executable repositories rather than as a single-turn code generator.
GPT-5.6 Luna is a fast, affordable GPT-5.6 model that brings strong capability at the lowest cost in the series.
GPT-5.6 Luna is a fast, affordable GPT-5.6 model that brings strong capability at the lowest cost in the series.
GPT-5.6 Sol is the flagship of OpenAI's GPT-5.6 series, its most capable model for long-horizon agentic work across coding, biology, and cybersecurity.
GPT-5.6 Sol is the flagship of OpenAI's GPT-5.6 series, its most capable model for long-horizon agentic work across coding, biology, and cybersecurity.
GPT-5.6 Terra is a balanced GPT-5.6 model for everyday work, with performance comparable to the previous generation at half the cost.
GPT-5.6 Terra is a balanced GPT-5.6 model for everyday work, with performance comparable to the previous generation at half the cost.
GPT-Realtime-2.1 updates GPT-Realtime-2 with improved alphanumeric recognition, silence and noise handling, and interruption behavior. It supports speech-to-speech interactions with configurable reasoning effort, instruction following, and tool use for complex voice-agent workflows.
Muse Spark 1.1 is strongest at agentic performance, tool use, and computer use. It does well on long-running tasks with 1M token context window, can delegate execution to sub-agents running in parallel, and is trained to use computer interfaces on desktop, mobile, or browser.
SpaceXAI's smartest model with frontier performance on coding, knowledge work, and STEM.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Claude Fable 5 is a Mythos-class model with robust safeguards. It can handle long-running, complex, and asynchronous tasks where previous models would have needed more frequent check-ins.
Gemini 3.1 Flash-Lite Image (Nano Banana 2 Lite) is Google's fastest image generation model enabling rapid creation and iteration.
Gemini Omni Flash (Preview) is a multimodal model designed for video, image, and text tasks. It is optimized for video generation, offering video output alongside text responses in a single model.
Sonnet 5 is an upgrade to Sonnet 4.6, with gains across agentic coding and professional work. It builds on the strengths of previous Sonnet models, bringing top-tier intelligence at Sonnet pricing for coding, agents, and everyday professional work at scale.
GLM-5.2-Fast-Preview is the high-speed variant of Zhipu AI's GLM-5.2, with 1M context and capabilities on par with the standard version. Inference-optimized to deliver 1.5–2× the output TPS, it fits latency-sensitive use cases such as real-time chat, multi-turn agents, and streaming code generation.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Seedance 2.0 Mini is a cost-effective multimodal video creation model, retaining the core functions and advantages of Seedance 2.0 at a lower price.
Fugu Ultra coordinates a deeper pool of expert agents to maximize answer quality on hard, high-stakes problems. Fugu Ultra focuses on maximizing performance, for a higher cost. It routes between one to three agents, depending on the problem.
GLM-5.2 delivers powerful coding capabilities, usable 1M-context support, and continued strengths in long-horizon tasks.
Kimi K2.7 Code HighSpeed is the high-speed version of Kimi K2.7 Code, the same model as Kimi K2.7 Code, but with an output speed of approximately 180 Tokens/s and up to 260 Tokens/s in short context scenarios, delivering a more extreme coding experience.
Kimi-K2.7-Code is a coding model from Moonshot AI. It has improved coding & agent performance over K2.6, more reasoning efficiency with less overthinking, and improved instruction following for long-horizon coding.
Hy-MT2-Lite is Tencent Cloud Hunyuan’s lightweight 1.8B translation model, combining efficient performance with an 8K-token context window and broad multilingual coverage. It supports translation across 33 languages and five ethnic Chinese and dialect variants, delivering strong results on benchmarks including FLORES-200 and WMT25. Designed for professional and real-world business use, it reliably follows instructions for structured, delimiter-preserving, context-aware, glossary-guided, and style-specific translation.
Hy-MT2-Plus is Tencent Cloud Hunyuan’s 7B translation model, offering an 8K-token context window and broad multilingual coverage. Optimized for translation across 33 languages and five ethnic Chinese and dialect variants, it delivers industry-leading performance on benchmarks including FLORES-200 and WMT25. It excels in professional domains and real-world business scenarios, with strong instruction-following for structured, delimiter-preserving, context-aware, glossary-guided, and style-specific translation.
A 550B parameter (55B active) open reasoning model from NVIDIA, built for long-running agent workflows. It uses a hybrid Mamba-Transformer MoE architecture and supports a 1M token context window.
Among the Qwen3.7 series, the cost-effective Plus model builds on its robust text capabilities while delivering a comprehensive upgrade to its vision‑language abilities, all while preserving its full‑stack agent‑level intelligence for coding, tool use, and productivity workflows.
MiniMax-M3 is a frontier-class foundation model that unites the three capabilities defining today's frontier: a 1M-token context window, frontier coding and agentic performance, and native multimodality — the first open-weight model to deliver all three in a single system.
Opus 4.8 is a focused upgrade to Opus 4.7 and is Anthropic's best generally available model for coding, agentic tasks, and enterprise workflows. It builds on the strengths of previous Opus models with stronger performance on complex, multi-step coding tasks. Anthropic recommends using it on long-horizon coding and agentic tasks. It is also stronger on professional work, including document drafting, data analysis, and presentations.
Opus 4.8 is a focused upgrade to Opus 4.7 and is Anthropic's best generally available model for coding, agentic tasks, and enterprise workflows. It builds on the strengths of previous Opus models with stronger performance on complex, multi-step coding tasks. Anthropic recommends using it on long-horizon coding and agentic tasks. It is also stronger on professional work, including document drafting, data analysis, and presentations.
Gemini 3.1 Flash Image is optimized for image understanding and generation and offers a balance of price and performance.
StepFun’s flagship multimodal reasoning model. Powered by a 198B-parameter / 11B-activation sparse MoE architecture, with native support for image and video understanding.
Qwen3.7 is a next‑generation flagship model designed for the agent‑centric era, with its core strengths lying in the breadth and depth of its agent‑level capabilities: it excels at programming, office and productivity tasks, and long‑term autonomous execution.
Hy-MT2-Pro is Tencent Cloud Hunyuan’s flagship translation model, featuring 30B-A3B parameters and an 8K-token context window. It delivers bidirectional translation across 33 core language pairs, plus five minority-language and dialect pairs. With industry-leading performance on FLORES-200 and WMT25, it excels in specialized domains and real-world business scenarios while supporting structured, contextual, terminology-aware, style-specific, and delimiter-preserving translation.
xAI's fast coding model trained specifically for agentic coding.
Google's latest model, highly optimized for coding proficiency and parallel agentic execution loops. Defaults to medium thinking effort for faster and more cost-efficient responses.
V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for.
V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for. V4.1 Pro generates higher-resolution images for when the idea deserves more room.
V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for. V4.1 Utility is designed for when restraint is the aesthetic choice, with flat lighting, front-facing composition, and simple, controlled scenes.
V4.1 is built on the same visual aesthetic and the same eye for what looks right, just pushed further across every dimension. The photorealism is more natural. The gradients are dreamier. Illustration styles are now possible that simply weren't before. And this is a model that reads a short prompts easier and creates something worth stopping for. V4.1 Utility is designed for when restraint is the aesthetic choice, with flat lighting, front-facing composition, and simple, controlled scenes.
Gemini 3.1 Flash Lite outperforms 2.5 Flash Lite on overall quality and lands close to 2.5 Flash performance across key capability areas. It is a workhorse model for high-volume use cases, with improvements across audio input/ASR, RAG snippet ranking, translation, data extraction, and code completion.
GPT Realtime 2 is our most capable realtime voice model. It supports speech-to-speech interactions with configurable reasoning effort, stronger instruction following, and more reliable tool use for complex voice-agent workflows.
GPT Realtime Whisper is a streaming speech-to-text model for applications that need low-latency transcript deltas from live audio. It is designed for realtime use cases where developers need to tune latency and accuracy. GPT Realtime Whisper is priced by audio duration rather than text tokens.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Mistral's frontier-class multimodal model optimized for agentic and coding use cases.
معلومات النموذج ومزوّدي التشغيل والأسعار.
GPT‑5.5 understands what you’re trying to do faster and can carry more of the work itself. It excels at writing and debugging code, researching online, analyzing data, creating documents and spreadsheets, operating software, and moving across tools until a task is finished. Instead of carefully managing every step, you can give GPT‑5.5 a messy, multi-part task and trust it to plan, use tools, check its work, navigate through ambiguity, and keep going.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Build real-time voice applications powered by Grok. Stream audio and text bidirectionally via WebSocket for voice assistants, phone agents, and interactive voice systems.
A native full-modal model supporting text, image, video, and audio understanding, with powerful Agent capabilities.
MiMo V2.5 Pro delivers significant improvements over its predecessor, MiMo-V2-Pro, in general agentic capabilities, complex software engineering, and long-horizon tasks. MiMo-V2.5-Pro is a 1.02T-parameter Mixture-of-Experts model with 42B active parameters, built on a hybrid-attention architecture with a 1M-token context window.
The Qwen3.6 35B-A3B native vision-language model is built on a hybrid architecture that integrates linear attention mechanisms with a sparse mixture-of-experts framework, achieving higher inference efficiency. Compared with the 3.5-35B-A3B, this model demonstrates significantly improved agentic coding capabilities, mathematical and code reasoning abilities, spatial intelligence, as well as object localization and object detection performance.
GPT Image 2 is OpenAI's state-of-the-art image generation model for fast, high-quality image generation and editing. It supports flexible image sizes and high-fidelity image inputs.
Kimi K2.6 demonstrates particularly strong performance in long-horizon coding tasks and produces professional-grade design with code and vision.
Compared with the previously released Qwen3-Max and Qwen3.6-Plus, this model features enhanced vibe coding abilities, more efficient coding agent execution, and significantly improved front-end development skills. Additionally, its long-tail knowledge retention has been further upgraded.
Best default for speed and quality, with stronger prompt following and more precise SVG structure.
Opus 4.7 builds on the coding and agentic strengths of Opus 4.6 with stronger performance on complex, multi-step tasks and more reliable agentic execution. It also brings improved performance on knowledge work, from drafting documents to building presentations and analyzing data.
Built with a unified multimodal audio-video joint generation architecture, Seedance 2.0 supports four input modalities: text, image, audio, and video. Compared with Version 1.5, Seedance 2.0 delivers a substantial leap in generation quality. It achieves a higher usability rate for complex interaction and motion scenes, with significant improvements in physical accuracy, visual realism, and controllability, making it well-suited for high-quality creation scenarios.
Seedance 2.0 Fast is a new-generation multimodal video creation model, inheriting the core functions and advantages of Seedance 2.0, with faster speed.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Gemma 4 31B is engineered to tackle the most demanding enterprise workloads and complex reasoning tasks. With an expansive 256K-token context window, the 31B model can effortlessly ingest entire codebases, and massive sets of images in a single prompt.
Google Gemma 4 26B A4B is a large language model from Google’s Gemma family. It is designed for high-quality text generation, reasoning, and conversational tasks. The model provides strong performance across a wide range of natural language applications, including chat, coding, and content generation.
The Qwen3.6 native vision-language Plus series models demonstrate exceptional performance on par with the current state-of-the-art models, with a significant improvement in overall results compared to the 3.5 series. The models have been markedly enhanced in code-related capabilities such as agentic coding, front-end programming, and Vibe coding, as well as in multi-modal general object recognition, OCR, and object localization.
Veo 3.1 Lite Preview is a high-efficiency, developer-first video model providing high-fidelity video generation, editing, and cinematic control. It leverages the state-of-the-art Veo 3.1 model to democratize professional-grade video AI by offering a scalable, programmable interface for creators and enterprises.
GLM-5V-Turbo is Z.AI’s first multimodal coding foundation model, built for vision-based coding tasks. It can natively process multimodal inputs such as images, video, and text, while also excelling at long-horizon planning, complex coding, and action execution. Deeply optimized for agent workflows, it works seamlessly with agents such as Claude Code and OpenClaw to complete the full loop of “understand the environment → plan actions → execute tasks”.
Trinity-Large-Thinking is a reasoning-optimized variant of Arcee AI's Trinity-Large family — a 398B-parameter sparse Mixture-of-Experts (MoE) model with approximately 13B active parameters per token. Built on Trinity-Large-Base and post-trained with extended chain-of-thought reasoning and agentic RL, Trinity-Large-Thinking delivers state-of-the-art performance on agentic benchmarks while maintaining strong general capabilities.
A high-performance edition designed for complex enterprise projects and SaaS integration.
M2.7 delivers outstanding performance in real-world software engineering, including end-to-end full project delivery, log analysis and bug troubleshooting, code security, machine learning, and more.
M2.7 Highspeed: Same performance, faster and more agile (output speed approximately 100 tps)
GPT-5.4 Mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.
GPT-5.4 Mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads.
GPT-5.4 Nano is designed for tasks where speed and cost matter most like classification, data extraction, ranking, and sub-agents.
Transcribe audio to text in 25 languages with batch and streaming modes.
Generate speech with 5 expressive voices, speech tags, and telephony codecs.
GLM 5 Turbo is a foundation model deeply optimized for the OpenClaw scenario. It has been specifically optimized for the core requirements of OpenClaw tasks since the training phase, enhancing key capabilities such as tool invocation, command following, timed and persistent tasks, and long-chain execution.
Grok 4.20 Beta is the newest flagship model from xAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherance, delivering consistently precise and truthful responses.
Grok 4.20 Beta is the newest flagship model from xAI with industry-leading speed and agentic tool calling capabilities. It combines the lowest hallucination rate on the market with strict prompt adherance, delivering consistently precise and truthful responses.
Multiple agents collaborate in parallel to perform deep research tasks.
NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. It delivers up to 7x higher throughput, providing fast, cost-efficient inference for agentic tasks. Additionally, a long context window gives the model long-term memory, preventing AI agents from losing focus on long, multi-step tasks and ensuring high-accuracy results. Fully open with weights, datasets, and recipes, Super allows easy customization and secure deployment anywhere.
Google’s first fully multimodal Embedding model that is capable of mapping text, image, video, audio, and PDFs and their interleaved combinations thereof into a single, unified vector space. Built on the Gemini architecture, it supports 100+ languages.
Multiple agents collaborate in parallel to perform deep research tasks.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Fish Audio S2 is the second-generation TTS model from Fish Audio. It's trained on over 10 million hours of audio across approximately 80 languages, and it introduces inline tag control: natural-language instructions embedded directly in your script at any position, giving you fine-grained direction over how speech is delivered at the word or phrase level.
Fish Audio S2 is the second-generation TTS model from Fish Audio. It's trained on over 10 million hours of audio across approximately 80 languages, and it introduces inline tag control: natural-language instructions embedded directly in your script at any position, giving you fine-grained direction over how speech is delivered at the word or phrase level.
GPT-5.4 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family. It's their most intelligent model yet for both general and agentic tasks.
GPT-5.4 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family. It's their most intelligent model yet for both general and agentic tasks.
GPT-5.4 Pro uses more compute to think harder and provide consistently better answers. It's designed to tackle tough problems.
Kling 3.0 delivers a major leap in character fidelity for motion-driven generation, with stable facial features across multi-angle and long-duration motion, accurate complex emotions from multi-image face references, identity preservation through partial occlusions (hats, hands, fans), and steady clarity as the camera zooms, pans, or tracks.
Turn spoken audio into accurate text — with timed segments — using Fish Audio’s ASR model. Send an audio file, get back the transcript, its duration, and timestamped segments.
Turn spoken audio into accurate text — with timed segments — using Fish Audio’s ASR model. Send an audio file, get back the transcript, its duration, and timestamped segments.
pplx-embed-v1 is a state-of-the-art text embedding model optimized for real-world, web-scale retrieval tasks.
pplx-embed-v1 is a state-of-the-art text embedding model optimized for real-world, web-scale retrieval tasks.
Gemini 3.1 Flash Image (Nano Banana 2) is optimized for image understanding and generation and offers a balance of price and performance.
A diffusion-based reasoning LLM that generates text via parallel refinement (not token-by-token), delivering real-time latency with ~1k tokens/sec plus 128K context and built-in tool/JSON support.
The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. Compared to the 3 series, these models deliver a leap forward in performance for both pure text and multimodal tasks, offering fast response times while balancing inference speed and overall performance.
GPT-Realtime-1.5 is our flagship audio model for voice agents and customer support.
Improved SWE and agentic capabilities, improved token efficiency and thinking, and expanded thinking levels.
Claude Sonnet 4.6 is the most capable Sonnet-class model yet, with frontier performance across coding, agents, and professional work. It excels at iterative development, complex codebase navigation, end-to-end project management with memory, polished document creation, and confident computer use for web QA and workflow automation.
The model delivers strong photorealism, including realistic skin rendering and natural textures, while avoiding common synthetic artifacts. It produces more distinctive lighting, composition, diverse subjects, contemporary styling, and carefully considered scene elements. For illustration, it generates original characters and forms with sophisticated and unexpected color combinations.
The model delivers strong photorealism, including realistic skin rendering and natural textures, while avoiding common synthetic artifacts. It produces more distinctive lighting, composition, diverse subjects, contemporary styling, and carefully considered scene elements. For illustration, it generates original characters and forms with sophisticated and unexpected color combinations.
The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In a variety of task evaluations, the 3.5 series consistently demonstrates performance on par with state-of-the-art leading models. Compared to the 3 series, these models show a leap forward in both pure-text and multimodal capabilities.
ByteDance-Seedream-5.0-lite is the latest image generation model released by BytePlus. For the first time, it introduces web-connected retrieval, enabling the model to fuse real-time online information to significantly improve the timeliness and relevance of generated images. The model’s reasoning and comprehension capabilities are further upgraded, allowing it to accurately interpret complex prompts and visual inputs. In addition, ByteDance-Seedream-5.0-lite delivers notable improvements in global knowledge coverage, reference consistency, and professional-grade scene generation, making it well suited for enterprise-level visual creation workflows.
GLM 5 is a frontier-class, general-purpose large language model optimized for complex systems engineering and long-horizon agentic tasks. It builds on the GLM 4.5 agent-centric lineage and is designed to support multi-step reasoning, math (including AIME-style benchmarks), advanced coding, and tool-augmented workflows, with long context support suitable for sophisticated agents and enterprise applications. Typical uses include autonomous agents for software engineering, data and systems troubleshooting, operations copilots, and high-end chat assistants that must break down complex tasks, call tools reliably, and reason over long sequences of instructions or documents.
MiniMax M2.5 is an agent-native frontier model trained explicitly to reason efficiently, decompose tasks optimally, and complete complex workflows under real-world time and cost constraints. It achieves task completion speeds comparable to or faster than leading proprietary frontier models by combining high inference throughput with reinforcement learning focused on token-efficient reasoning and better decision-making in agentic scaffolds. M2.5 is well-suited for production agents handling full-stack software projects, research and analysis workflows, long-horizon planning, and multi-tool orchestration, delivering intelligence for organizations that need both high-quality reasoning and predictable, low latency.
M2.5 highspeed: Same performance, faster and more agile (output speed approximately 100 tps)
Opus 4.6 is the world’s best model for coding and professional work, built to power agents that take on whole categories of real-world work. It excels across the entire SDLC, breaking through on hard problems, identifying complex bugs, and demonstrating deeper codebase understanding. It also delivers a step-change in knowledge work, with near-production-ready documents, presentations, and spreadsheets on the first pass.
GPT-5.3-Codex advances both the frontier coding performance of GPT‑5.2-Codex and the reasoning and professional knowledge capabilities of GPT‑5.2, together in one model, which is also 25% faster. This enables it to take on long-running tasks that involve research, tool use, and complex execution.
GPT-5.3-Codex advances both the frontier coding performance of GPT‑5.2-Codex and the reasoning and professional knowledge capabilities of GPT‑5.2, together in one model, which is also 25% faster. This enables it to take on long-running tasks that involve research, tool use, and complex execution.
Build upon an All-in-One product framework, the Kling 3.0 model series supports full multimodal input and output spanning text, images, audio, and video, bringing the understanding, generation, and editing of video together in one streamlined AI workflow. The models integrate multiple tasks, including text-to-video, image-to-video, reference-to-video, and in-video editing, into a single, native multimodal architecture, enabling the models to follow complex narrative logic, deliver precise shot control, and maintain strong prompt adherence.
Build upon an All-in-One product framework, the Kling 3.0 model series supports full multimodal input and output spanning text, images, audio, and video, bringing the understanding, generation, and editing of video together in one streamlined AI workflow. The models integrate multiple tasks, including text-to-video, image-to-video, reference-to-video, and in-video editing, into a single, native multimodal architecture, enabling the models to follow complex narrative logic, deliver precise shot control, and maintain strong prompt adherence.
Step 3.5 Flash is an open-source reasoning model by StepFun with 196B total parameters (11B active) using Mixture of Experts. It features a 256K context window, deep reasoning, tool calling, and agentic capabilities, achieving 97.3 on AIME 2025 and 74.4% on SWE-bench Verified.
State-of-the-art video generation across quality, cost, and latency. Grok Imagine is x.AI's most powerful video-audio generative model yet. Bring an image to life, start from a simple text prompt, or even refine a complex cinematic sequence.
Generate high-quality images from text prompts with xAI's imagine API.
kimi-k2.5 is Kimi's most versatile model to date, featuring a native multimodal architecture that supports both visual and text input, thinking and non-thinking modes, and dialogue and agent tasks.
Compared with the snapshot as of September 23, 2025, the Qwen-3 series Max model in this release achieves an effective integration of thinking and non-thinking modes, resulting in a comprehensive and substantial improvement in the model’s overall performance. In thinking mode, the model simultaneously supports web search, web information extraction, and a code interpreter tool, enabling it to tackle more complex and challenging problems with greater accuracy by leveraging external tools while engaging in slow, deliberative reasoning. This version is based on a snapshot taken on January 23, 2026.
GLM-4.7-Flash balances high performance with efficiency, making it the perfect lightweight deployment option. Beyond coding, it is also recommended for creative writing, translation, long-context tasks, and roleplay.
GLM-4.7-Flash balances high performance with efficiency, making it the perfect lightweight deployment option.
FLUX.2 [klein] is Black Forest Labs' fastest image model yet - it unifies image generation and editing in a single, compact model. Delivering state-of-the-art quality with end-to-end inference in less than a second. Enabling interactive workflows, real-time previews, and latency-critical applications.
FLUX.2 [klein] is Black Forest Labs' fastest image model yet - it unifies image generation and editing in a single, compact model. Delivering state-of-the-art quality with end-to-end inference in less than a second. Enabling interactive workflows, real-time previews, and latency-critical applications.
Optimized for general-purpose and multilingual retrieval quality. All embeddings created with the 4 series are compatible with each other.
The best general-purpose and multilingual retrieval quality. All embeddings created with the 4 series are compatible with each other.
Optimized for latency and cost. All embeddings created with the 4 series are compatible with each other.
MiniMax 2.1 is MiniMax's latest model, optimized specifically for robustness in coding, tool use, instruction following, and long-horizon planning.
MiniMax-M2.1-lightning is a faster version of MiniMax-M2.1, offering the same performance but with significantly higher throughput (output speed ~100 TPS, MiniMax-M2 output speed ~60 TPS).
GLM-4.7 is Z.ai’s latest flagship model, with major upgrades focused on two key areas: stronger coding capabilities and more stable multi-step reasoning and execution.
GPT‑5.2-Codex is a version of GPT‑5.2 further optimized for agentic coding in Codex, including improvements on long-horizon work through context compaction, stronger performance on large code changes like refactors and migrations, improved performance in Windows environments, and significantly stronger cybersecurity capabilities.
Kling 2.6 introduces a groundbreaking "Native Audio" capability, enabling the generation of complete videos in a single go, including natural voice, action sound effects, and environmental ambient sounds, providing an immersive "what you see if what you hear" experience.
Google's most intelligent model built for speed, combining frontier intelligence with superior search and grounding.
FLUX.2 [max] offers image generation and image editing with the highest quality available. It delivers state-of-the-art image generation and advanced image editing with exceptional realism, precision, and consistency. Built for professional use, FLUX.2 [max] produces production-ready outputs for marketing teams, creatives, filmmakers, and creators around the world.
GPT Image 1.5 is OpenAI's latest image generation model, with better instruction following and adherence to prompts.
ByteDance's Seedance 1.5 Pro is a professional video model using V2A native generation for integrated, synced audio-visual output, enhancing efficiency of professional video creation.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
NVIDIA Nemotron 3 Nano is an open reasoning model optimized for fast, cost-efficient inference. Built with a hybrid MoE and Mamba architecture and trained on NVIDIA-curated synthetic reasoning data, it delivers strong multi-step reasoning with stable latency and predictable performance for agentic and production workloads.
A light version of Rerank 4 Pro, this is a multilingual model that allows for re-ranking English and non-english documents and semi-structured data (JSON). This model is better suited for low latency and high throughput use-cases than its pro variant.
A multilingual model that allows for re-ranking English and non-english documents and semi-structured data (JSON). This model is better suited for state-of-the-art quality and complex use-cases than its fast variant.
GPT-5.2 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family. It's their most intelligent model yet for both general and agentic tasks.
Version of GPT-5.2 that produces smarter and more precise responses.
GPT-5.2 is OpenAI's best general-purpose model, part of the GPT-5 flagship model family. It's their most intelligent model yet for both general and agentic tasks.
An enterprise-grade text model that excels at using tools to explore codebases, editing multiple files, and powering software engineering agents.
Our open source model that excels at using tools to explore codebases, editing multiple files, and powering software engineering agents.
Kling 2.6 introduces a groundbreaking "Native Audio" capability, enabling the generation of complete videos in a single go, including natural voice, action sound effects, and environmental ambient sounds, providing an immersive "what you see if what you hear" experience.
Kling 2.6 introduces a groundbreaking "Native Audio" capability, enabling the generation of complete videos in a single go, including natural voice, action sound effects, and environmental ambient sounds, providing an immersive "what you see if what you hear" experience.
Seedream 4.5 is the latest in-house image generation model developed by ByteDance. Compared with Seedream 4.0, it delivers comprehensive improvements—especially in editing consistency, including better preservation of subject details, lighting, and color tone. It also enhances portrait refinement and small-text rendering. The model’s multi-image composition capabilities have been significantly strengthened, and both reasoning performance and visual aesthetics continue to advance, enabling more accurate and artistically expressive image generation.
GPT OSS Safeguard 120B is OpenAI’s open-weight safety model for content moderation and guardrail enforcement. It helps teams evaluate content against custom safety policies and build scalable, adaptable protections for production AI applications.
Ministral 3 14B is the largest model in the Ministral 3 family, offering state-of-the-art capabilities and performance comparable to its larger Mistral Small 3.2 24B counterpart. Optimized for local deployment, it delivers high performance across diverse hardware, including local setups.
Mistral Large 3 2512 is Mistral’s most capable model to date. It has a sparse mixture-of-experts architecture with 41B active parameters (675B total).
Nova 2 Lite is a fast, cost-effective reasoning model for everyday workloads that can process text, images, and videos to generate text.
DeepSeek‑V3.2 from DeepSeek harmonizes high computational efficiency with superior reasoning and agent performance. It builds on three main techniques: DeepSeek Sparse Attention for long‑context efficiency, a scalable reinforcement learning framework, and a large‑scale agentic task synthesis pipeline. This model excels at long-context reasoning and agentic tasks, efficiently handling extended inputs while maintaining strong accuracy. Its sparse attention design enables it to process complex, multi-step workflows without excessive compute costs. Overall, DeepSeek‑V3.2 targets long‑context reasoning, tool‑using agents, and efficient deployment in production environments.
DeepSeek‑V3.2 from DeepSeek harmonizes high computational efficiency with superior reasoning and agent performance. It builds on three main techniques: DeepSeek Sparse Attention for long‑context efficiency, a scalable reinforcement learning framework, and a large‑scale agentic task synthesis pipeline. This model excels at long-context reasoning and agentic tasks, efficiently handling extended inputs while maintaining strong accuracy. Its sparse attention design enables it to process complex, multi-step workflows without excessive compute costs. Overall, DeepSeek‑V3.2 targets long‑context reasoning, tool‑using agents, and efficient deployment in production environments.
FLUX.2 is a completely new base model trained for visual intelligence, not just pixel generation, setting a new standard for both image generation and image editing. With FLUX.2 models you can expect the highest quality, higher resolutions (up to 4MP), and new capabilities like multi-ref images. FLUX.2 [flex] supports customizable image generation and editing with adjustable steps and guidance. It's better at typography and text rendering. It supports up to 10 reference images (up to 14 MP total input). This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
FLUX.2 is a completely new base model trained for visual intelligence, not just pixel generation, setting a new standard for both image generation and image editing. With FLUX.2 models you can expect the highest quality, higher resolutions (up to 4MP), and new capabilities like multi-ref images. FLUX.2 [pro] supports generation, editing, and multiple reference images (up to 9 MP total input). This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
Claude Opus 4.5 is Anthropic’s latest model in the Opus series, meant for demanding reasoning tasks and complex problem solving. This model has improvements in general intelligence and vision compared to previous iterations. In addition, it is suited for difficult coding tasks and agentic workflows, especially those with computer use and tool use, and can effectively handle context usage and external memory files.
GPT‑5.1-Codex-Max is purpose-built for agentic coding.
معلومات النموذج ومزوّدي التشغيل والأسعار.
معلومات النموذج ومزوّدي التشغيل والأسعار.
The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B).
GPT-5.1 Codex mini is a smaller, faster, and cheaper version of GPT-5.1 Codex.
An upgraded version of GPT-5 that adapts thinking time more precisely to the question to spend more time on complex questions and respond more quickly to simpler tasks.
An upgraded version of GPT-5 that adapts thinking time more precisely to the question to spend more time on complex questions and respond more quickly to simpler tasks.
GPT-5.1-Codex is a version of GPT-5.1 optimized for agentic coding tasks in Codex or similar environments.
KAT-Coder-Pro V1 is KwaiKAT's most advanced agentic coding model in the KwaiKAT series. Designed specifically for agentic coding tasks, it excels in real-world software engineering scenarios, achieving a remarkable 73.4% solve rate on the SWE-Bench Verified benchmark. KAT-Coder-Pro V1 delivers top-tier coding performance and has been rigorously tested by thousands of in-house engineers. The model has been optimized for tool-use capability, multi-turn interaction, instruction following, generalization and comprehensive capabilities through a multi-stage training process, including mid-training, supervised fine-tuning (SFT), reinforcement fine-tuning (RFT), and scalable agentic RL.
Kimi K2 Thinking is an advanced open-source thinking model by Moonshot AI. It can execute up to 200 – 300 sequential tool calls without human interference, reasoning coherently across hundreds of steps to solve complex problems. Built as a thinking agent, it reasons step by step while using tools, achieving state-of-the-art performance on Humanity's Last Exam (HLE), BrowseComp, and other benchmarks, with major gains in reasoning, agentic search, coding, writing, and general capabilities.
OpenAI's first open weight reasoning model specifically trained for safety classification tasks. Fine-tuned from GPT-OSS, this model helps classify text content based on customizable policies, enabling bring-your-own-policy Trust & Safety AI where your own taxonomy, definitions, and thresholds guide classification decisions.
The model is an auto-regressive vision language model that uses an optimized transformer architecture. The model enables multi-image reasoning and video understanding, along with strong document intelligence, visual Q&A and summarization capabilities.
MiniMax-M2 redefines efficiency for agents. It is a compact, fast, and cost-effective MoE model (230 billion total parameters with 10 billion active parameters) built for elite performance in coding and agentic tasks, all while maintaining powerful general intelligence.
Seedance 1.0 Pro Fast delivers top performance at an unbeatable price, balancing quality, speed, and cost. Built on Seedance 1.0 Pro’s core strengths, it’s faster and more cost-efficient for creators.
Fish Audio S1 is trained on over 2 million hours of audio with online RLHF (GRPO). It achieves 0.8% WER and 0.4% CER on Seed TTS Eval. S1 supports open-domain emotion, tone, and special effect markers.
Fish Audio S1 is trained on over 2 million hours of audio with online RLHF (GRPO). It achieves 0.8% WER and 0.4% CER on Seed TTS Eval. S1 supports open-domain emotion, tone, and special effect markers.
Claude Haiku 4.5 matches Sonnet 4's performance on coding, computer use, and agent tasks at substantially lower cost and faster speeds. It delivers near-frontier performance and Claude’s unique character at a price point that works for scaled sub-agent deployments, free tier products, and intelligence-sensitive applications with budget constraints.
Veo 3.1 is Google's state-of-the-art model for generating high-fidelity, 8-second 720p, 1080p or 4k videos featuring stunning realism and natively generated audio.
Veo 3.1 Fast is a specialized, high-speed variant of Google DeepMind’s Veo 3.1 text-to-video model, optimized for rapid generation of 8-second, high-fidelity videos. It is designed to create cinematic, 1080p, or 720p content with improved prompt adherence and native audio, making it ideal for creating quick, high-quality video clips, social media content, and ad creatives.
GPT-Realtime mini is capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.
Interfaze is an AI model built on a new architecture that merges specialized DNN/CNN models with LLMs for developer tasks that require deterministic output and high consistency like OCR, scraping, classification, STT and more.
A cost-efficient version of GPT Image 1. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.
GPT-5 pro uses more compute to think harder and provide consistently better answers. Since GPT-5 pro is designed to tackle tough problems, some requests may take several minutes to finish.
As the latest iteration in the GLM series, GLM-4.6 achieves comprehensive enhancements across multiple domains, including real-world coding, long-context processing, reasoning, searching, writing, and agentic applications.
Claude Sonnet 4.5 is the newest model in the Sonnet series, offering improvements and updates over Sonnet 4.
معلومات النموذج ومزوّدي التشغيل والأسعار.
Kling 2.5 Turbo is a major update to the AI video generation model focused on significantly improving speed, video quality, temporal stability, and creative control for creators, making professional-grade AI-generated video faster, more coherent, and easier to direct from text prompts.
Kling 2.5 Turbo is a major update to the AI video generation model focused on significantly improving speed, video quality, temporal stability, and creative control for creators, making professional-grade AI-generated video faster, more coherent, and easier to direct from text prompts.
The Qwen 3 series Max model has undergone specialized upgrades in agent programming and tool invocation compared to the preview version. The officially released model this time has achieved state-of-the-art (SOTA) performance in its field and is better suited to meet the demands of agents operating in more complex scenarios.
The Qwen3 series VL models has been comprehensively upgraded in areas such as visual coding and spatial perception. Its visual perception and recognition capabilities have significantly improved, supporting the understanding of ultra-long videos, and its OCR functionality has undergone a major enhancement.
The Qwen3 series VL models has been comprehensively upgraded in areas such as visual coding and spatial perception. Its visual perception and recognition capabilities have significantly improved, supporting the understanding of ultra-long videos, and its OCR functionality has undergone a major enhancement.
Qwen3 series VL models feature significantly enhanced multimodal reasoning capabilities, with a particular focus on optimizing the model for STEM and mathematical reasoning. Visual perception and recognition abilities have been comprehensively improved, and OCR capabilities have undergone a major upgrade.
Qwen3 series VL models feature significantly enhanced multimodal reasoning capabilities, with a particular focus on optimizing the model for STEM and mathematical reasoning. Visual perception and recognition abilities have been comprehensively improved, and OCR capabilities have undergone a major upgrade.
DeepSeek-V3.1-Terminus delivers more stable & reliable outputs across benchmarks compared to the previous version and addresses user feedback (i.e. language consistency and agent upgrades).
GPT-5-Codex is a version of GPT-5 optimized for agentic coding tasks in Codex or similar environments.
A new generation of open-source, non-thinking mode model powered by Qwen3. This version demonstrates superior Chinese text understanding, augmented logical reasoning, and enhanced capabilities in text generation tasks over the previous iteration (Qwen3-235B-A22B-Instruct-2507).
A new generation of Qwen3-based open-source thinking mode models. This version offers improved instruction following and streamlined summary responses over the previous iteration (Qwen3-235B-A22B-Thinking-2507).
Seedream 4.0 is a SOTA multimodal image creation model built on leading architecture. It breaks through the boundaries of traditional text-to-image models by natively supporting text, single-image, and multi-image inputs. Users can freely combine text and images to achieve diverse creative modes within a single model—such as multi-image blending, image editing, and sequentially batch image generation, featuring subject consistency, making image creation more free and controllable.
Qwen3-Max-Preview shows substantial gains over the 2.5 series in overall capability, with significant enhancements in Chinese-English text understanding, complex instruction following, handling of subjective open-ended tasks, multilingual ability, and tool invocation; model knowledge hallucinations are reduced.
Bytedance Seed 1.8 features stronger multimodal understanding and agent capabilities. The model delivers superior performance across a wide range of complex real-world tasks, helping enterprises create greater value.
Nano Banana Pro (Gemini 3 Pro Image) builds on Nano Banana's generation capabilities into a new era of studio-quality, functional design to help you create and edit high-fidelity, production-ready visuals with unparalleled precision and control. Improvements include enhanced world knowledge and reasoning, dynamic text and translation, and studio level controls.
ByteDance's new multimodal deep-thinking model, supporting both text and visual inputs with enhanced reasoning capabilities.
Nano Banana (Gemini 2.5 Flash Image) is Google's first fully hybrid reasoning model, letting developers turn thinking on or off and set thinking budgets to balance quality, cost, and latency. Upgraded for rapid creative workflows, it can generate interleaved text and images and supports conversational, multi‑turn image editing in natural language. It’s also locale‑aware, enabling culturally and linguistically appropriate image generation for audiences worldwide.
DeepSeek-V3.1 is post-trained on the top of DeepSeek-V3.1-Base, which is built upon the original V3 base checkpoint through a two-phase long context extension approach, following the methodology outlined in the original DeepSeek-V3 report. We have expanded our dataset by collecting additional long documents and substantially extending both training phases. The 32K extension phase has been increased 10-fold to 630B tokens, while the 128K extension phase has been extended by 3.3x to 209B tokens. Additionally, DeepSeek-V3.1 is trained using the UE8M0 FP8 scale data format to ensure compatibility with microscaling data formats.
NVIDIA-Nemotron-Nano-9B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and designed as a unified model for both reasoning and non-reasoning tasks. It responds to user queries and tasks by first generating a reasoning trace and then concluding with a final response. The model's reasoning capabilities can be controlled via a system prompt. If the user prefers the model to provide its final answer without intermediate reasoning traces, it can be configured to do so.\
Built on the GLM-4.5-Air base model, GLM-4.5V inherits proven techniques from GLM-4.1V-Thinking while achieving effective scaling through a powerful 106B-parameter MoE architecture.
A generalist reranker optimized for quality with instruction-following and multilingual support.
A generalist reranker optimized for both latency and quality with instruction-following and multilingual support.
GPT-5 is OpenAI's flagship language model that excels at complex reasoning, broad real-world knowledge, code-intensive, and multi-step agentic tasks.
GPT-5 is OpenAI's flagship language model that excels at complex reasoning, broad real-world knowledge, code-intensive, and multi-step agentic tasks.
GPT-5 mini is a cost optimized model that excels at reasoning/chat tasks. It offers an optimal balance between speed, cost, and capability.
GPT-5 mini is a cost optimized model that excels at reasoning/chat tasks. It offers an optimal balance between speed, cost, and capability.
GPT-5 nano is a high throughput model that excels at simple instruction or classification tasks.
Extremely capable general-purpose LLM with strong, controllable reasoning capabilities
A compact, open-weight language model optimized for low-latency and resource-constrained environments, including local and edge deployments.
Efficient coding specialist balancing performance with cost-effectiveness for daily development tasks while maintaining strong tool integration capabilities.
Veo 3 Fast is a quicker and more cost effective version of Veo 3, allowing developers to create videos with sound while maintaining high quality and optimizing for speed and business use cases. Veo 3 Fast offers both text-to-video and image-to-video modalities.
GLM-4.5 and GLM-4.5-Air are our latest flagship models, purpose-built as foundational models for agent-oriented applications. Both leverage a Mixture-of-Experts (MoE) architecture. GLM-4.5 has a total parameter count of 355B with 32B active parameters per forward pass, while GLM-4.5-Air adopts a more streamlined design with 106B total parameters and 12B active parameters.
GLM-4.5 and GLM-4.5-Air are our latest flagship models, purpose-built as foundational models for agent-oriented applications. Both leverage a Mixture-of-Experts (MoE) architecture. GLM-4.5 has a total parameter count of 355B with 32B active parameters per forward pass, while GLM-4.5-Air adopts a more streamlined design with 106B total parameters and 12B active parameters.
Powered by Qwen3 this is a powerful Coding Agent that excels in tool calling and environment interaction to achieve autonomous programming. It combines outstanding coding proficiency with versatile general-purpose abilities.
Qwen3-Coder-480B-A35B-Instruct is a cutting-edge open coding model from Qwen, matching Claude Sonnet’s performance in agentic programming, browser automation, and core development tasks.
Qwen3-Coder-Next is an open-weight language model built specifically for coding, with strong performance on large-scale software engineering and agentic coding benchmarks. It uses a hybrid Mixture-of-Experts architecture to offer high capability at relatively modest active parameter counts, improving efficiency for real-world deployments. The model is trained on diverse code and natural language data so it can handle tasks like code generation, refactoring, debugging, repository-level reasoning, and technical explanation across multiple programming languages. It is also optimized for tool use and function calling, making it suitable as the core of coding agents that interact with shells, editors, issue trackers, and other developer tools.
Kimi K2 is a state-of-the-art mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters. Trained with the Muon optimizer, Kimi K2 achieves exceptional performance across frontier knowledge, reasoning, and coding tasks while being meticulously optimized for agentic capabilities.
Morph offers a specialized AI model that applies code changes suggested by frontier models (like Claude or GPT-4o) to your existing code files FAST - 4500+ tokens/second. It acts as the final step in the AI coding workflow. Supports 16k input tokens and 16k output tokens.
Morph offers a specialized AI model that applies code changes suggested by frontier models (like Claude or GPT-4o) to your existing code files FAST - 2500+ tokens/second. It acts as the final step in the AI coding workflow. Supports 16k input tokens and 16k output tokens.
Gemini 2.5 Flash-Lite is a balanced, low-latency model with configurable thinking budgets and tool connectivity (e.g., Google Search grounding and code execution). It supports multimodal input and offers a 1M-token context window.
A video generation model that supports multi-shot storytelling. It excels in semantic understanding and instruction following, producing smooth, detailed, and cinematic 1080P HD videos.
The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers.
The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B).
The Qwen3 Embedding model series is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. Building upon the dense foundational models of the Qwen3 series, it provides a comprehensive range of text embeddings and reranking models in various sizes (0.6B, 4B, and 8B).
FLUX.1 Kontext creates images from text prompts with unique capabilities for character consistency and advanced editing. It also edits images using simple text prompts. No complex workflows or fine-tuning needed. This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
FLUX.1 Kontext creates images from text prompts with unique capabilities for character consistency and advanced editing. It also edits images using simple text prompts. No complex workflows or fine-tuning needed. This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
Code embedding model that can embed code databases and repositories to power coding assistants.
Claude Opus 4 is Anthropic's most powerful model yet and the state-of-the-art coding model. It delivers sustained performance on long-running tasks that require focused effort and thousands of steps, significantly expanding what AI agents can solve. Claude Opus 4 is ideal for powering frontier agent products and features.
Claude Sonnet 4 balances impressive performance for coding with the right speed and cost for high-volume use cases: Coding: Handle everyday development tasks with enhanced performance-power code reviews, bug fixes, API integrations, and feature development with immediate feedback loops.
State-of-the-art embedding model with excellent performance across English, multilingual and code tasks.
Veo 3 is designed to handle a range of video generation tasks, from cinematic narratives to dynamic character animations. With Veo 3, you can create more immersive experiences by not only generating stunning visuals, but also audio like dialogue and sound effects.
Voyage AI's embedding model optimized for general-purpose and multilingual retrieval quality.
Voyage AI's embedding model optimized for latency and cost.
GPT 4.1 mini provides a balance between intelligence, speed, and cost that makes it an attractive model for many use cases.
GPT 4.1 mini provides a balance between intelligence, speed, and cost that makes it an attractive model for many use cases.
Mistral Medium 3 delivers frontier performance while being an order of magnitude less expensive. For instance, the model performs at or above 90% of Claude Sonnet 3.7 on benchmarks across the board at a significantly lower cost.
Qwen3-32B is a world-class model with comparable quality to DeepSeek R1 while outperforming GPT-4.1 and Claude Sonnet 3.7. It excels in code-gen, tool-calling, and advanced reasoning, making it an exceptional model for a wide range of production use cases.
Qwen3-235B-A22B-Instruct-2507 is the updated version of the Qwen3-235B-A22B non-thinking mode, featuring Significant improvements in general capabilities, including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage.
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support
Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support
OpenAI's o3 is their most powerful reasoning model, setting new state-of-the-art benchmarks in coding, math, science, and visual perception. It excels at complex queries requiring multi-faceted analysis, with particular strength in analyzing images, charts, and graphics.
OpenAI's o3 is their most powerful reasoning model, setting new state-of-the-art benchmarks in coding, math, science, and visual perception. It excels at complex queries requiring multi-faceted analysis, with particular strength in analyzing images, charts, and graphics.
OpenAI's o4-mini delivers fast, cost-efficient reasoning with exceptional performance for its size, particularly excelling in math (best-performing on AIME benchmarks), coding, and visual tasks.
OpenAI's o4-mini delivers fast, cost-efficient reasoning with exceptional performance for its size, particularly excelling in math (best-performing on AIME benchmarks), coding, and visual tasks.
A model that allows for text, images, or mixed content to be classified or turned into embeddings.
GPT 4.1 is OpenAI's flagship model for complex tasks. It is well suited for problem solving across domains.
GPT 4.1 is OpenAI's flagship model for complex tasks. It is well suited for problem solving across domains.
GPT-4.1 nano is the fastest, most cost-effective GPT 4.1 model.
GPT-4.1 nano is the fastest, most cost-effective GPT 4.1 model.
As a general purpose LLM, Llama 4 Maverick contains 17 billion active parameters, 128 experts, and 400 billion total parameters, offering high quality at a lower price compared to Llama 3.3 70B.
Llama 4 Scout is the best multimodal model in the world in its class and is more powerful than our Llama 3 models, while fitting in a single H100 GPU. Additionally, Llama 4 Scout supports an industry-leading context window of up to 10M tokens.
GPT Image 1 is OpenAI's new state-of-the-art image generation model. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.
Gemini 2.5 Flash is a thinking model that offers great, well-rounded capabilities. It is designed to offer a balance between price and performance with multimodal support and a 1M token context window.
Gemini 2.5 Pro is our most advanced reasoning Gemini model, capable of solving complex problems. Gemini 2.5 Pro can comprehend vast datasets and challenging problems from different information sources, including text, audio, images, video, and even entire code repositories.
Command A is Cohere's most performant model to date, excelling at tool use, agents, retrieval augmented generation (RAG), and multilingual use cases. Command A has a context length of 256K, only requires two GPUs to run, and has 150% higher throughput compared to Command R+ 08-2024.
Mercury Coder Small is ideal for code generation, debugging, and refactoring tasks with minimal latency.
Perplexity's lightweight offering with search grounding, quicker and cheaper than Sonar Pro.
Perplexity's premier offering with search grounding, supporting advanced queries and follow-ups.
A premium reasoning-focused model that outputs Chain of Thought (CoT) in responses, providing comprehensive explanations with enhanced search capabilities and multiple search queries per request.
o3-mini is OpenAI's most recent small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini.
DeepSeek-R1 provides customers a state-of-the-art reasoning model, optimized for general reasoning tasks, math, science, and code generation.
Voyage AI's embedding model with the best general-purpose and multilingual retrieval quality.
Where performance meets efficiency. This model supports high-performance conversational AI designed for content creation, enterprise applications, and research, offering advanced language understanding capabilities, including text summarization, classification, sentiment analysis, and code generation.
o1 is OpenAI's flagship reasoning model, designed for complex problems that require deep thinking. It provides strong reasoning capabilities with improved accuracy for complex multi-step tasks.
Voyage AI's embedding model optimized for code retrieval.
A very low cost multimodal model that is lightning fast for processing image, video, and text inputs.
A text-only model that delivers the lowest latency responses at very low cost.
A highly capable multimodal model with the best combination of accuracy, speed, and cost for a wide range of tasks.
A model that allows for re-ranking English Language documents and semi-structured data (JSON).
FLUX1.1 [pro] Ultra delivers ultra-fast, ultra high-resolution image creation - with more pixels in every picture. Generate varying aspect ratios from text, at 4MP resolution fast. This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
V3 introduced major advances in photorealism and text rendering. It was the first Recraft model to generate mid-size text accurately and, as of 2025, is the only model capable of placing text at specific positions in an image.
A compact, efficient model for on-device tasks like smart assistants and local analytics, offering low-latency performance.
A more powerful model with faster, memory-efficient inference, ideal for complex workflows and demanding edge applications.
FLUX1.1 [pro] is the standard for text-to-image generation with fast, reliable and consistently stunning results. This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
A state-of-the-art inpainting model, enabling editing and expansion of real and generated images given a text description and a binary mask. This provider gives the option to change the moderation level for inputs and outputs. The control is under safety tolerance and is by default 2 on a range from 0 (more strict) through 6 (more permissive).
Mistral Small is the ideal choice for simple tasks that one can do in bulk - like Classification, Customer Support, or Text Generation. It offers excellent performance at an affordable price point.
A 12B model with image understanding capabilities in addition to text.
Lightning-fast image generation
English-focused text embedding model optimized for code and English language tasks.
An update to Meta Llama 3 70B Instruct that includes an expanded 128K context length, multilinguality and improved reasoning capabilities.
An update to Meta Llama 3 8B Instruct that includes an expanded 128K context length, multilinguality and improved reasoning capabilities.
GPT-4o mini from OpenAI is their most advanced and cost-efficient small model. It is multi-modal (accepting text or image inputs and outputting text) and has higher intelligence than gpt-3.5-turbo but is just as fast.
GPT-4o mini from OpenAI is their most advanced and cost-efficient small model. It is multi-modal (accepting text or image inputs and outputting text) and has higher intelligence than gpt-3.5-turbo but is just as fast.
12B model trained jointly by Mistral AI and NVIDIA, it significantly outperforms existing models smaller or similar in size.
Voyage AI's embedding model optimized for finance retrieval and RAG.
Mistral's cutting-edge language model for coding released end of July 2025, Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation.
GPT-4o from OpenAI has broad general knowledge and domain expertise allowing it to follow complex instructions in natural language and solve difficult problems accurately. It matches GPT-4 Turbo performance with a faster and cheaper API.
GPT-4o from OpenAI has broad general knowledge and domain expertise allowing it to follow complex instructions in natural language and solve difficult problems accurately. It matches GPT-4 Turbo performance with a faster and cheaper API.
Amazon Titan Text Embeddings V2 is a light weight, efficient multilingual embedding model supporting 1024, 512, and 256 dimensions.
Voyage AI's embedding model optimized for legal retrieval and RAG.
gpt-4-turbo from OpenAI has broad general knowledge and domain expertise allowing it to follow complex instructions in natural language and solve difficult problems accurately. It has a knowledge cutoff of April 2023 and a 128,000 token context window.
Claude 3 Haiku is Anthropic's fastest, most compact model for near-instant responsiveness. It answers simple queries and requests with speed. Customers will be able to build seamless AI experiences that mimic human interactions. Claude 3 Haiku can process images and return text outputs, and features a 200K context window.
GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio. It offers improvements to word error rate and better language recognition and accuracy compared to original Whisper models. Use it for more accurate transcripts.
GPT-4o Transcribe is a speech-to-text model that uses GPT-4o to transcribe audio. It offers improvements to word error rate and better language recognition and accuracy compared to original Whisper models. Use it for more accurate transcripts.
Recraft V2 is an image generation model released in March 2024 and the first model trained from scratch by Recraft. With 20 billion parameters, it was a breakthrough in human anatomical accuracy and the first to support brand consistency and brand color inputs. It also introduced vector image generation (SVG output), as well as minimalistic icon and illustration styles.
Multilingual text embedding model optimized for cross-lingual tasks across many languages.
OpenAI's most capable embedding model for both english and non-english tasks.
OpenAI's improved, more performant version of their ada embedding model.
Voyage AI's embedding model optimized for code retrieval (17% better than alternatives). This is the previous generation of code embeddings models.
General-purpose text embedding model for semantic search, similarity, clustering, and RAG workflows.
TTS is a model that converts text to natural sounding spoken text.
TTS is a model that converts text to natural sounding spoken text. The tts-1-hd model is optimized for high quality text-to-speech use cases.
OpenAI's most capable and cost effective model in the GPT-3.5 family optimized for chat purposes, but also works well for traditional completions tasks.
OpenAI's legacy text embedding model.
Whisper is a general-purpose speech recognition model, trained on a large dataset of diverse audio. You can also use it as a multitask model to perform multilingual speech recognition as well as speech translation and language identification.
Muse Spark 1.3 is Meta’s multimodal reasoning model for long-horizon agentic and coding workflows. With a 1M-token context window, reliable tool calling, higher first-attempt accuracy, and native understanding of video, images, and documents, it helps developers build capable coding agents and AI development workflows with fewer unnecessary turns and cleaner output.
Muse Spark 1.3 Contributor is Meta’s cost-optimized API option for agentic coding workflows. It offers a 1M-token context window, dependable tool calling, and multimodal perception at $0.10 per million input tokens and $0.20 per million output tokens, with usage permitted to improve Meta’s products.