Skip to navigation
AI Gateway

Supported models

Model ids the AI Gateway accepts, and what each one is for
View as Markdown

Pass any id on this page as model on the SDK call. Chat models share one API, so switching models means changing only the model string.

Chat models

Anthropic Claude

claude-opus-5-5

Newest and most capable Claude. Hardest reasoning and long agentic work.

claude-opus-5

Previous Opus. Reasons before it answers, so give it a larger max_tokens.

claude-sonnet-5

Newest balanced Claude. Also reasons before it answers.

claude-opus-4-8

Longer reasoning and agentic workflows.

claude-sonnet-4-6

Workhorse with balanced cost and quality. A good default.

claude-haiku-4-5

Fast and cheap. Tool-use loops and classification.

OpenAI GPT

gpt-5-6-sol

Most capable GPT. Frontier reasoning and agentic coding.

gpt-5-6-terra

Balanced GPT for production. Code, extraction, and agents.

gpt-5-6-luna

Fast and cheap GPT. Classification, summaries, and routing.

gpt-5-5

Previous GPT generation. General chat and reasoning.

gpt-5-4

Previous GPT generation. General chat and reasoning.

Amazon Nova

nova-pro

Balanced general use with low latency.

nova-lite

Cheaper and faster for high-volume calls.

nova-micro

Lowest-cost text tier.

Mistral and Qwen

mistral-large-3

Mistral flagship. Strong general reasoning and coding.

pixtral-large

Mistral vision-language. Accepts image and text input.

qwen3-coder

Qwen coding specialist. Code generation and refactoring.

qwen3-vl

Qwen vision-language. Accepts image and text input.

Embedding models

titan-embed-text-v2

Text embeddings for RAG and vector search.

Decision models

Decision models answer typed questions about a state you describe. They don’t chat. Call them on the decisions endpoint.

jev-1-13-0

TypeSafe jev judgment model. Returns a probability, an option, or a score for each question.

Next steps