Available Models

Most capable and efficient frontier model with 1M context, native computer use, and thinking mode

Premium GPT-5.4 with maximum compute for the hardest problems

High intelligence with medium speed. Multimodal with vision, function calling, and structured outputs

Frontier model with 400K context and adaptive reasoning

Strongest mini model for coding, computer use, and subagents with GPT-5.4 capabilities

Cost-optimized reasoning and chat

Fastest and most affordable GPT-5.4 model for high-throughput tasks

Uses more compute for consistently better answers

Industry-leading agentic coding model. 400K context, reasoning, tool use, and complex execution

Advanced reasoning model for complex tasks

Fast reasoning model optimized for STEM

Latest reasoning model with improved performance

Efficient reasoning model for STEM tasks

Open-weight 20B model (Apache 2.0), similar performance to o3-mini. Available on testnet for developer testing.

Open-weight 120B model (Apache 2.0), flagship open model. Available on testnet for developer testing.

Fastest and most efficient Claude, near-frontier intelligence

Best balance of intelligence, speed, and cost

Latest Anthropic flagship with enhanced reasoning and creativity

Latest flagship Claude with extended 64k output, vision, and advanced reasoning

Latest Gemini with improved thinking, token efficiency, and agentic capabilities. Optimized for software engineering (requires new SDK)

Flagship frontier model for high-precision multimodal reasoning

Frontier-class performance with Pro-level intelligence at Flash speed and pricing. Includes thinking mode (requires new SDK)

State-of-the-art for reasoning, coding, and mathematics

Fast and efficient Gemini model with vision support

Ultra-fast and lightweight Gemini 3.1 model with thinking mode for high-throughput tasks

Most economical Gemini model - ultra-fast and lightweight (requires new SDK)

DeepSeek V3.2 non-thinking mode, excellent for chat and coding

DeepSeek V3.2 thinking mode for complex reasoning tasks

Z.AI's flagship foundation model with 200K context. Strong reasoning and agentic capabilities

Optimized GLM-5 variant with faster inference

MiniMax's flagship reasoning model with recursive self-improvement. Great value for complex tasks (~60 tps)

OpenAI's open-weight 120B model hosted free by NVIDIA. Apache 2.0 license, great for experimentation

OpenAI's open-weight 20B model hosted free by NVIDIA. Fast and efficient for simpler tasks

Moonshot's flagship MoE model (1T params) hosted by NVIDIA. Vision and agentic capabilities

NVIDIA's flagship 253B reasoning model. Strong on math, coding, and instruction following

NVIDIA MoE model (12B active params) with thinking mode. Fast and capable reasoning

NVIDIA Nemotron 49B with thinking mode. Good balance of speed and reasoning quality

DeepSeek's latest V3.2 MoE model hosted free by NVIDIA. Same quality, zero cost

Mistral's flagship 675B model hosted free by NVIDIA. Largest Mistral model ever released

Qwen's 480B MoE coding model (35B active) hosted by NVIDIA. Optimized for code generation

Mistral's 123B coding-focused model hosted free by NVIDIA. Strong code and instruction following

Zhipu AI's GLM-4.7 with thinking mode hosted by NVIDIA. Unique Chinese AI lab model

Meta's Llama 4 Maverick MoE (17B x 128 experts) hosted free by NVIDIA