GPT-5.4 is our frontier model for complex professional work. Reasoning.effort supports: none (default), low, medium, high and xhigh. Features a 1.05M ...

GPT-5.4 is our frontier model for complex professional work. Reasoning.effort supports: none (default), low, medium, high and xhigh. Features a 1.05M ...

Gemini 3.1 Flash-Lite Preview is Google's most cost-efficient model, optimized for high-volume agentic tasks, translation, and simple data processing....

Claude Sonnet 4.6 is Anthropic's most capable Sonnet model, released February 2026. Features near-Opus-level intelligence at Sonnet pricing, with a 1M...

Gemini 3.1 Pro Preview is Google's most advanced reasoning model, released February 2026. It uses extended thinking/chain-of-thought reasoning to work...

Claude Opus 4.6 is Anthropic's most capable model to date, released February 2026. Building on the intelligence of Opus 4.5, it brings new levels of r...

Gemini 3 Flash Preview is Google's latest fast and efficient AI model optimized for quick response times while maintaining high quality. This preview ...

GPT-5.2 is our best general-purpose model, part of the GPT-5 flagship model family. Our most intelligent model yet for both general and agentic tasks,...

Tough problems that may take longer to solve but require harder thinking

GPT-5.2 Chat is a continuously updated version of GPT-5.2 optimized for conversational interactions. It receives regular updates with the latest impro...

GPT Image 1.5 is OpenAI's state-of-the-art image generation model with better instruction following, 4× faster generation, and cheaper image tokens th...

Claude Opus 4.5 is Anthropic's flagship model released November 2025, representing the highest level of intelligence and capability. Features extended...

Gemini 3 Pro Image is Google's native image generation model with state-of-the-art reasoning capabilities. It is the best model for complex and multi-...

Gemini 3 Pro Preview is Google's latest experimental AI model with advanced reasoning, coding, and multimodal capabilities. This preview version offer...

A frontier multimodal model optimized specifically for high-performance agentic tool calling.

A frontier multimodal model optimized for high-performance agentic tool calling with reasoning capabilities.

GPT-5.1 is an enhanced version of GPT-5 with improved performance and capabilities. It features the same 400K context window and advanced tool calling...

Kimi K2 Thinking is a powerful open-source AI model from Moonshot AI designed for complex, step-by-step reasoning and long-horizon agentic tasks. It e...

Our fastest model. Intelligence at blazing speeds. Multilingual and vision capabilities. 8,192 max output tokens. Training data cut-off: October 2024....

Our fastest model. Intelligence at blazing speeds. Multilingual and vision capabilities. 8,192 max output tokens. Training data cut-off: October 2024....

Most capable GPT-5 model with extended thinking capabilities

Best-in-class coding and agentic model with hours-long autonomous operation capabilities. Supports extended thinking, context awareness, parallel tool...

Best-in-class coding and agentic model with hours-long autonomous operation capabilities. Supports extended thinking, context awareness, parallel tool...

Qwen3 VL 235B A22B Instruct is a powerful, open-weight multimodal model from Alibaba Cloud that excels at both language and vision tasks. It integrate...

DeepSeek-V3.1 Terminus is an update to DeepSeek V3.1 that maintains the model's original capabilities while addressing issues reported by users, inclu...

DeepSeek-V3.2-Exp is an experimental model introducing the groundbreaking DeepSeek Sparse Attention (DSA) mechanism for enhanced long-context processi...

Grok 4 Fast is xAI's latest advancement in cost-efficient reasoning models. Built on xAI’s learnings from Grok 4, Grok 4 Fast delivers frontier-level ...

Enhanced version of Kimi K2 with doubled context window (256k tokens) and significantly improved coding capabilities, especially for frontend developm...

Grok 4 Fast is xAI's latest advancement in cost-efficient reasoning models. Built on xAI’s learnings from Grok 4, Grok 4 Fast delivers frontier-level ...

GPT-5 is OpenAI's most advanced language model, featuring enhanced reasoning capabilities with 80% fewer factual errors than o3. It supports a 400K to...

GPT-5 Mini delivers GPT-5-level performance at a fraction of the cost and latency. With the same 400K context window and advanced capabilities includi...

GPT-5 Nano is the smallest and fastest model in the GPT-5 family, designed for ultra-low latency applications. Despite its compact size, it maintains ...

Our most capable model with the highest level of intelligence and capability. Supports extended thinking, multilingual capabilities, and vision proces...

Our most capable model with the highest level of intelligence and capability. Supports extended thinking, multilingual capabilities, and vision proces...

This streamlined model maintains impressive performance and efficiency, featuring the following key enhancements: (a) Significant Performance among op...

Qwen3-235B-A22B-Thinking-2507 is the Qwen3's new model with scaling the thinking capability of Qwen3-235B-A22B, improving both the quality and depth o...

Qwen3-Coder-480B-A35B-Instruct is the Qwen3's most agentic code model, featuring significant performance on agentic coding, agentic browser-use and ot...

Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improv...

DeepSeek-TNG-R1T2-Chimera is the second-generation Chimera model from TNG Tech. It is a 671 B-parameter mixture-of-experts text-generation model assem...

Gemini 2.5 Flash is Google's state-of-the-art workhorse model, specifically designed for advanced reasoning, coding, mathematics, and scientific tasks...

Gemini 2.5 Pro is Google’s state-of-the-art AI model designed for advanced reasoning, coding, mathematics, and scientific tasks. It employs “thinking”...

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. B...

Our previous flagship model with very high intelligence and capability. Supports extended thinking, multilingual capabilities, and vision processing. ...

High-performance model with high intelligence and balanced performance. Supports extended thinking, multilingual capabilities, and vision processing. ...

Qwen3-32B is a 32.8 billion parameter language model that uniquely supports seamless switching between thinking mode for complex reasoning tasks and n...

GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. ...

GPT-4.1 Mini is a mid-sized model delivering performance competitive with GPT-4o at substantially lower latency and cost. It retains a 1 million token...

For tasks that demand low latency, GPT-4.1 nano is the fastest and cheapest model in the GPT-4.1 series. It delivers exceptional performance at a smal...

GPT Image 1 is OpenAI's image generation model that turns text and image inputs into high-fidelity images. It offers strong instruction following and ...

ERNIE-4.5-21B-A3B-Thinking is a text-based Mixture of Experts (MoE) post-training model featuring 21B total parameters with 3B active parameters per t...

High-performance model with toggleable extended thinking for complex reasoning tasks. Combines high intelligence with the ability to think through pro...

Kimi K2.5 is Moonshot AI's flagship agentic model and a new SOTA open model. Built on Kimi K2 with continued pretraining over approximately 15T mixed ...

Fast and accurate web-grounded chat model with real-time search capabilities. Ideal for general queries requiring up-to-date information from the web.

Advanced web-grounded chat model with enhanced search quality and 200K context window. Best for complex queries requiring comprehensive web research.

Web-grounded reasoning model that thinks step-by-step before responding. Combines search capabilities with logical reasoning for accurate, well-reason...

Advanced reasoning model with 128K context window designed for complex, multi-step queries. Provides in-depth analysis with web-grounded research and ...

Specialized research model that conducts comprehensive multi-query searches with citation tracking and reasoning tokens. Automatically determines sear...

DeepSeek-R1-Distill-Llama-70B is a 70-billion parameter model created by distilling the reasoning capabilities of DeepSeek's flagship R1 model (671B p...

DeepSeek-Reasoner (DeepSeek-V3.1 Thinking Mode) is designed for advanced reasoning, mathematical problem-solving, and complex coding tasks. It uses ch...

Reasoning model with extended thinking capabilities

Efficient reasoning model

GPT-5 is OpenAI's most advanced language model, featuring enhanced reasoning capabilities with 80% fewer factual errors than o3. It supports a 400K to...

GPT-5 Mini delivers GPT-5-level performance at a fraction of the cost and latency. With the same 400K context window and advanced capabilities includi...

GPT-5 Nano is the smallest and fastest model in the GPT-5 family, designed for ultra-low latency applications. Despite its compact size, it maintains ...

Most capable GPT-5 model with extended thinking capabilities

Specialized model for code generation and analysis

GPT-5.1 is an enhanced version of GPT-5 with improved performance and capabilities. It features the same 400K context window and advanced tool calling...

Specialized model for code generation and analysis, based on GPT-5.1

Compact specialized model for code generation and analysis, based on GPT-5.1

GPT-5.1 Chat is a continuously updated version of GPT-5.1 optimized for conversational interactions. It receives regular updates with the latest impro...

Latest version of Codex Mini, a compact specialized model for code generation and analysis

Llama 4 instruction-tuned MoE (17B, 16 experts) for fast, high-quality chat, tool use, and multilingual reasoning with balanced latency and cost.

Llama 4 instruction-tuned MoE (17B, 128 experts) targeting tougher reasoning and long-form tasks, trading more compute for higher response diversity a...

Meta’s latest safety/guardrail model for prompt and output moderation, aligning conversations to policy via classification and constrained generation.

Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 bill...

Qwen3-Next-80B-A3B-Instruct is a causal language model that is instruction-optimized for chat and agent applications. It features a Mixture-of-Experts...

DeepSeek-V3.1 (deepseek-chat) is a powerful generalist model with 671B parameters, offering exceptional performance at an economical price. It achieve...

GLM-4.7 is Zhipu AI's flagship coding model with major upgrades in advanced coding capabilities, multi-step reasoning, and agentic orchestration. Feat...

Experimental version of Gemini 2.0 Flash with native image generation capabilities. Features multimodal input and output support including text and im...

Llama-3.3-70B-Versatile is Meta's advanced multilingual large language model, optimized for a wide range of natural language processing tasks. With 70...

The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama...

Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini mode...

Our previous intelligent model with high level of intelligence and capability. Fast latency with multilingual and vision capabilities, but no extended...

Our fastest model. Intelligence at blazing speeds. Multilingual and vision capabilities. 8,192 max output tokens. Training data cut-off: July 2024. AP...

86M parameter multilingual prompt safety classifier based on mDeBERTa-base, detecting prompt injections and jailbreaks across 8+ languages with advers...

22M parameter lightweight prompt safety classifier based on DeBERTa-xsmall, offering 75% reduced latency for detecting prompt injections and jailbreak...

GPT-5 Chat Latest is a continuously updated version of GPT-5 optimized for conversational interactions. It receives regular updates with the latest im...

Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language mo...

Speedy and economical reasoning model that excels at agentic coding. Features function calling, structured outputs, and reasoning capabilities.

OpenAI ChatGPT 4o is continually updated by OpenAI to point to the current version of GPT-4o used by ChatGPT. It therefore differs slightly from the A...

Mistral Large 2.1

Meta's latest class of models, Llama 3.1, launched with a variety of sizes and configurations. The 8B instruct-tuned version is particularly fast and ...

Optimized version of Llama 3.1 8B Instruct with 128K context window, designed for high-speed inference in multilingual chat and dialogue use cases wit...

GPT-4o mini is OpenAI's newest model after GPT-4 Omni, supporting both text and image inputs with text outputs. As their most advanced small model, i...

The Mistral-Nemo-Instruct-2407 Large Language Model (LLM) is an instruct fine-tuned version of the Mistral-Nemo-Base-2407. Trained jointly by Mistral ...

As the latest iteration in the GLM series, GLM-4.6 achieves comprehensive enhancements across multiple domains, including real-world coding, long-cont...

Latest and greatest flagship model, offering unparalleled performance in natural language, math and reasoning - the perfect jack of all trades. Featur...

Compact 8B general-purpose model offering efficient inference for chat, coding, and RAG workflows on limited compute.

Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini mode...

o3 is a well-rounded and powerful model across domains. It sets a new standard for math, science, coding, and visual reasoning tasks. It also excels a...

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 c...

o4-mini is our latest small o-series model. It's optimized for fast, effective reasoning with exceptionally efficient performance in coding and visual...

gpt-oss-120b is our most powerful open-weight model, which fits into a single H100 GPU (117B parameters with 5.1B active parameters). Features permiss...

gpt-oss-20b is our medium-sized open-weight model for low latency, local, or specialized use-cases (21B parameters with 3.6B active parameters). Featu...

Excels at enterprise use cases like data extraction, coding, and text summarization. Possesses deep domain knowledge in finance, healthcare, law, and ...

Lightweight model that thinks before responding. Fast, smart, and great for logic-based tasks that do not require deep domain knowledge. Features func...

Hermes 2 Pro is an upgraded, retrained version of Nous Hermes 2, consisting of an updated and cleaned version of the OpenHermes 2.5 Dataset, as well a...

GPT-4o ("o" for "omni") is OpenAI's latest AI model, supporting both text and image inputs with text outputs. It maintains the intelligence level of G...

Claude 3 Haiku is Anthropic's fastest and most compact model. Designed for near-instant responsiveness and seamless AI experiences that mimic human in...

Mistral-Small-3.2-24B-Instruct-2506 is an updated 24B parameter model from Mistral optimized for instruction following, repetition reduction, and impr...

o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini. o3-mini supports key develop...