Models & protocols
Model IDs, tiers, and which endpoint serves which family.
Model IDs
Send one of these IDs as model. IDs are case-sensitive. You can also fetch the live list from GET /v1/models, or browse rates on the Models page.
| Model ID | Family | Tier | Context |
|---|---|---|---|
| claude-fable-5-2 | Claude | Frontier | 262K |
| claude-fable-5-1 | Claude | Frontier | 262K |
| claude-fable-5 | Claude | Frontier | 262K |
| claude-mythos-5 | Claude | Frontier | 262K |
| claude-opus-5.5 | Claude | Frontier | 262K |
| claude-opus-5 | Claude | Frontier | 262K |
| claude-opus-4.8 | Claude | Frontier | 262K |
| claude-opus-4.7 | Claude | Frontier | 262K |
| claude-opus-4.6 | Claude | Frontier | 262K |
| claude-opus-4.5 | Claude | Frontier | 262K |
| claude-opus-4.1 | Claude | Frontier | 262K |
| claude-opus-4 | Claude | Frontier | 262K |
| claude-sonnet-5.5 | Claude | Balanced | 131K |
| claude-sonnet-5.1 | Claude | Balanced | 131K |
| claude-sonnet-5 | Claude | Balanced | 131K |
| claude-sonnet-4.6 | Claude | Balanced | 131K |
| claude-sonnet-4.5 | Claude | Balanced | 131K |
| claude-sonnet-4 | Claude | Balanced | 131K |
| claude-haiku-5.5 | Claude | Speed | 131K |
| claude-haiku-4.5 | Claude | Speed | 131K |
| claude-haiku-3.5 | Claude | Speed | 131K |
| gpt-astra-6 | ChatGPT | Frontier | 262K |
| gemini-argon-4 | Gemini | Frontier | 131K |
| gemini-4-carbon | Gemini | Balanced | 1.05M |
| gemini-3-pro | Gemini | Frontier | 131K |
| gemini-2.5-pro | Gemini | Frontier | 131K |
| gemini-2.5-flash | Gemini | Balanced | 131K |
| gemini-2.5-flash-lite | Gemini | Speed | 131K |
| gpt-5 | ChatGPT | Frontier | 262K |
| gpt-5-mini | ChatGPT | Balanced | 131K |
| gpt-4.1 | ChatGPT | Balanced | 131K |
| gpt-4.1-mini | ChatGPT | Speed | 131K |
| gpt-4o | ChatGPT | Balanced | 131K |
| gpt-4o-mini | ChatGPT | Speed | 131K |
| deepseek-v4 | DeepSeek | Frontier | 131K |
| deepseek-r1 | DeepSeek | Frontier | 131K |
| deepseek-v3 | DeepSeek | Balanced | 131K |
| grok-4 | Grok | Frontier | 131K |
| grok-3 | Grok | Balanced | 131K |
| kimi-k2-thinking | Kimi | Frontier | 131K |
| kimi-k2 | Kimi | Frontier | 131K |
| qwen-3-max | Qwen | Frontier | 131K |
| qwen-3-coder | Qwen | Balanced | 131K |
| mistral-large-3 | Mistral | Frontier | 131K |
| codestral | Mistral | Balanced | 131K |
| mistral-small | Mistral | Speed | 131K |
| llama-4-maverick | Llama | Frontier | 131K |
| llama-4-scout | Llama | Balanced | 131K |
| gemini-4-hydrogen | Gemini | Balanced | 1.05M |
| muse-ember | Llama | Speed | 1.05M |
| hy4 | Hunyuan | Speed | 1.05M |
| glm-5.3-flashx-legal | GLM | Speed | 1.05M |
| gpt-6.1-terra-pro | ChatGPT | Balanced | 1.05M |
| gpt-6.1-terra | ChatGPT | Balanced | 1.05M |
| kimi-k3.1-pro | Kimi | Balanced | 1.05M |
| minimax-m3.1-flash | MiniMax | Speed | 1.05M |
| deepseek-v4.1-flash-1017 | DeepSeek | Speed | 1.05M |
| muse-spark-open | Llama | Balanced | 1.05M |
| nemotron-3.5-super-vl | Nemotron | Speed | 1.05M |
| muse-wonder | Llama | Frontier | 1.05M |
| mistral-large-4-0 | Mistral | Balanced | 1.05M |
| gpt-6.1-sol-pro | ChatGPT | Balanced | 1.05M |
| gpt-6.1-sol | ChatGPT | Balanced | 1.05M |
| glm-5.3-prime | GLM | Balanced | 1M |
| qwen3.8-max-prime | Qwen | Balanced | 1M |
| command-a-plus | Command | Balanced | 192K |
| gpt-6-luna-pro | ChatGPT | Speed | 1.05M |
| gpt-6-luna | ChatGPT | Speed | 1.05M |
| gpt-6-sol-pro | ChatGPT | Balanced | 1.05M |
| gpt-6-sol | ChatGPT | Balanced | 1.05M |
| mimo-v2.6-pro-ultraspeed | MiMo | Balanced | 1.05M |
| mimo-v2.6-flash | MiMo | Speed | 1.05M |
| mimo-v2.6-pro | MiMo | Speed | 1.05M |
| grok-4.7 | Grok | Balanced | 500K |
| qwen3.8-omni-flash | Qwen | Speed | 1M |
| glm-5.3-flashx | GLM | Speed | 1.05M |
| deepseek-v4.1-flash | DeepSeek | Speed | 1.05M |
| gpt-6-astra-pro | ChatGPT | Frontier | 1.05M |
| qwen3.8-max-0902 | Qwen | Balanced | 1M |
| muse-spark-1.3-contributor | Llama | Speed | 1.05M |
| muse-spark-1.3 | Llama | Balanced | 1.05M |
| gemini-3.8-flash | Gemini | Balanced | 1.05M |
| hy4-preview | Hunyuan | Balanced | 1.05M |
| qwen3.8-flash | Qwen | Speed | 1M |
| glm-5.3-flash | GLM | Speed | 1.05M |
| muse-spark-1.2-contributor | Llama | Speed | 1.05M |
| deepseek-v4-flash-vision-exp | DeepSeek | Speed | 1.05M |
| hy-mt2-1.8b | Hunyuan | Speed | 8K |
| hy-mt2-30b-a3b | Hunyuan | Speed | 8K |
| hy-mt2-7b | Hunyuan | Speed | 8K |
| glm-5.3 | GLM | Speed | 1.05M |
| qwen3.8-27b | Qwen | Speed | 1.05M |
| gemini-3.7-flash | Gemini | Balanced | 1.05M |
| qwen3.8-2.4t-a95b | Qwen | Balanced | 1.05M |
| deepseek-v4-pro-0813 | DeepSeek | Balanced | 1.05M |
| grok-4.6 | Grok | Balanced | 500K |
| nemotron-3.5-lightning | Nemotron | Speed | 262K |
| muse-glimmer-30b | Llama | Speed | 131K |
| muse-spark-1.2 | Llama | Balanced | 1.05M |
| deepseek-v4-flash-0731 | DeepSeek | Speed | 1.05M |
| qwen3.7-flash | Qwen | Speed | 1M |
| gemini-3.6-flash | Gemini | Balanced | 1.05M |
| gemini-3.5-flash-lite | Gemini | Balanced | 1.05M |
| kimi-k3 | Kimi | Balanced | 1.05M |
| muse-spark-1.1 | Llama | Balanced | 1.05M |
| gpt-5.6-luna-pro | ChatGPT | Speed | 1.05M |
| gpt-5.6-luna | ChatGPT | Speed | 1.05M |
| gpt-5.6-terra-pro | ChatGPT | Balanced | 1.05M |
| gpt-5.6-terra | ChatGPT | Balanced | 1.05M |
| gpt-5.6-sol-pro | ChatGPT | Frontier | 1.05M |
| gpt-5.6-sol | ChatGPT | Frontier | 1.05M |
| grok-4.5 | Grok | Balanced | 500K |
| hy3 | Hunyuan | Speed | 262K |
| glm-5.2 | GLM | Balanced | 1.05M |
| kimi-k2.7-code | Kimi | Balanced | 262K |
| nemotron-3.5-content-safety | Nemotron | Speed | 131K |
| nemotron-3-ultra-550b-a55b | Nemotron | Balanced | 262K |
| qwen3.7-plus | Qwen | Speed | 1M |
| minimax-m3 | MiniMax | Speed | 1.05M |
| qwen3.7-max | Qwen | Balanced | 1M |
| grok-build-0.1 | Grok | Balanced | 256K |
| gemini-3.5-flash | Gemini | Balanced | 1.05M |
| gemini-3.1-flash-lite | Gemini | Balanced | 1.05M |
| grok-4.3 | Grok | Balanced | 1M |
| mistral-medium-3-5 | Mistral | Balanced | 262K |
| qwen3.5-plus-20260420 | Qwen | Balanced | 1M |
| qwen3.6-flash | Qwen | Speed | 1M |
| qwen3.6-35b-a3b | Qwen | Speed | 262K |
| qwen3.6-27b | Qwen | Balanced | 262K |
| gpt-5.5-pro | ChatGPT | Frontier | 1.05M |
| gpt-5.5 | ChatGPT | Frontier | 1.05M |
| deepseek-v4-flash | DeepSeek | Speed | 1.05M |
| hy3-preview | Hunyuan | Speed | 262K |
| mimo-v2.5-pro | MiMo | Speed | 1.02M |
| mimo-v2.5 | MiMo | Balanced | 1M |
| kimi-k2.6 | Kimi | Balanced | 262K |
| glm-5.1 | GLM | Balanced | 204K |
| gemma-4-26b-a4b-it | Gemini | Speed | 262K |
| gemma-4-31b-it | Gemini | Speed | 262K |
| qwen3.6-plus | Qwen | Balanced | 1M |
| glm-5v-turbo | GLM | Balanced | 202K |
| grok-4.20-multi-agent | Grok | Balanced | 2M |
| grok-4.20 | Grok | Balanced | 2M |
| minimax-m2.7 | MiniMax | Speed | 196K |
| gpt-5.4-nano | ChatGPT | Speed | 400K |
| gpt-5.4-mini | ChatGPT | Balanced | 400K |
| glm-5-turbo | GLM | Balanced | 202K |
| nemotron-3-super-120b-a12b | Nemotron | Speed | 262K |
| qwen3.5-9b | Qwen | Speed | 262K |
| gpt-5.4-pro | ChatGPT | Frontier | 1.05M |
| gpt-5.4 | ChatGPT | Frontier | 1.05M |
| gemini-3.1-flash-lite-preview | Gemini | Balanced | 1.05M |
| qwen3.5-35b-a3b | Qwen | Speed | 262K |
| qwen3.5-27b | Qwen | Balanced | 262K |
| qwen3.5-122b-a10b | Qwen | Balanced | 262K |
| qwen3.5-flash-02-23 | Qwen | Speed | 1M |
| gemini-3.1-pro-preview-customtools | Gemini | Balanced | 1.05M |
| gpt-5.3-codex | ChatGPT | Balanced | 400K |
| gemini-3.1-pro-preview | Gemini | Balanced | 1.05M |
| qwen3.5-plus-02-15 | Qwen | Balanced | 1M |
| qwen3.5-397b-a17b | Qwen | Balanced | 262K |
| minimax-m2.5 | MiniMax | Speed | 200K |
| glm-5 | GLM | Balanced | 204K |
| qwen3-coder-next | Qwen | Speed | 262K |
| kimi-k2.5 | Kimi | Balanced | 262K |
| glm-4.7-flash | GLM | Speed | 200K |
| gpt-5.2-codex | ChatGPT | Balanced | 400K |
| glm-4.7 | GLM | Balanced | 204K |
| gemini-3-flash-preview | Gemini | Balanced | 1.05M |
| nemotron-3-nano-30b-a3b | Nemotron | Speed | 262K |
| gpt-5.2-chat | ChatGPT | Balanced | 128K |
| gpt-5.2-pro | ChatGPT | Frontier | 400K |
| gpt-5.2 | ChatGPT | Balanced | 400K |
| devstral-2512 | Mistral | Balanced | 262K |
| glm-4.6v | GLM | Speed | 131K |
| gpt-5.1-codex-max | ChatGPT | Balanced | 400K |
| ministral-14b-2512 | Mistral | Speed | 262K |
| ministral-8b-2512 | Mistral | Speed | 262K |
| ministral-3b-2512 | Mistral | Speed | 131K |
| deepseek-v3.2 | DeepSeek | Speed | 163K |
| gpt-5.1 | ChatGPT | Balanced | 400K |
| gpt-5.1-codex | ChatGPT | Balanced | 400K |
| gpt-5.1-codex-mini | ChatGPT | Balanced | 400K |
| sonar-pro-search | Sonar | Frontier | 200K |
| voxtral-small-24b-2507 | Mistral | Speed | 32K |
| gpt-oss-safeguard-20b | ChatGPT | Speed | 131K |
| qwen3-vl-8b-instruct | Qwen | Speed | 262K |
| qwen3-vl-30b-a3b-thinking | Qwen | Speed | 262K |
| qwen3-vl-30b-a3b-instruct | Qwen | Speed | 262K |
| gpt-5-pro | ChatGPT | Frontier | 400K |
| glm-4.6 | GLM | Balanced | 204K |
| deepseek-v3.2-exp | DeepSeek | Speed | 163K |
| qwen3-vl-235b-a22b-instruct | Qwen | Speed | 262K |
| deepseek-v3.1-terminus | DeepSeek | Speed | 163K |
| qwen3-coder-flash | Qwen | Speed | 1M |
| qwen3-next-80b-a3b-instruct | Qwen | Speed | 262K |
| kimi-k2-0905 | Kimi | Balanced | 262K |
| deepseek-chat-v3.1 | DeepSeek | Speed | 163K |
| mistral-medium-3.1 | Mistral | Balanced | 131K |
| glm-4.5v | GLM | Balanced | 65K |
| gpt-5-nano | ChatGPT | Speed | 400K |
| gpt-oss-120b | ChatGPT | Speed | 131K |
| gpt-oss-20b | ChatGPT | Speed | 131K |
| qwen3-coder-30b-a3b-instruct | Qwen | Speed | 262K |
| qwen3-30b-a3b-instruct-2507 | Qwen | Speed | 262K |
| glm-4.5 | GLM | Balanced | 131K |
| glm-4.5-air | GLM | Speed | 131K |
| qwen3-235b-a22b-thinking-2507 | Qwen | Balanced | 128K |
| qwen3-235b-a22b-2507 | Qwen | Speed | 262K |
| hunyuan-a13b-instruct | Hunyuan | Speed | 131K |
| mistral-small-3.2-24b-instruct | Mistral | Speed | 256K |
| o3-pro | ChatGPT | Frontier | 200K |
| gemini-2.5-pro-preview | Gemini | Balanced | 1.05M |
| mistral-medium-3 | Mistral | Balanced | 131K |
| llama-guard-4-12b | Llama | Speed | 163K |
| qwen3-30b-a3b | Qwen | Speed | 40K |
| qwen3-14b | Qwen | Speed | 40K |
| qwen3-32b | Qwen | Speed | 131K |
| o3 | ChatGPT | Balanced | 200K |
| gpt-4.1-nano | ChatGPT | Speed | 1.05M |
| deepseek-chat-v3-0324 | DeepSeek | Speed | 163K |
| mistral-small-3.1-24b-instruct | Mistral | Speed | 128K |
| gemma-3-4b-it | Gemini | Speed | 131K |
| gemma-3-12b-it | Gemini | Speed | 131K |
| command-a | Command | Balanced | 256K |
| gemma-3-27b-it | Gemini | Speed | 131K |
| sonar-reasoning-pro | Sonar | Balanced | 128K |
| sonar-pro | Sonar | Frontier | 200K |
| sonar-deep-research | Sonar | Balanced | 128K |
| mistral-saba | Mistral | Speed | 32K |
| qwen2.5-vl-72b-instruct | Qwen | Speed | 128K |
| qwen-plus | Qwen | Speed | 1M |
| mistral-small-24b-instruct-2501 | Mistral | Speed | 32K |
| sonar | Sonar | Speed | 127K |
| command-r7b-12-2024 | Command | Speed | 128K |
| llama-3.3-70b-instruct | Llama | Speed | 131K |
| gpt-4o-2024-11-20 | ChatGPT | Balanced | 128K |
| mistral-large-2407 | Mistral | Balanced | 131K |
| qwen-2.5-coder-32b-instruct | Qwen | Speed | 32K |
| llama-3.2-1b-instruct | Llama | Speed | 60K |
| llama-3.2-3b-instruct | Llama | Speed | 131K |
| qwen-2.5-72b-instruct | Qwen | Speed | 32K |
| command-r-08-2024 | Command | Speed | 128K |
| command-r-plus-08-2024 | Command | Balanced | 128K |
| gpt-4o-2024-08-06 | ChatGPT | Balanced | 128K |
| llama-3.1-70b-instruct | Llama | Speed | 131K |
| llama-3.1-8b-instruct | Llama | Speed | 131K |
| mistral-nemo | Mistral | Speed | 131K |
| gpt-4o-mini-2024-07-18 | ChatGPT | Speed | 128K |
| gemma-2-27b-it | Gemini | Speed | 8K |
| gpt-4o-2024-05-13 | ChatGPT | Frontier | 128K |
| mixtral-8x22b-instruct | Mistral | Balanced | 65K |
| mistral-large | Mistral | Balanced | 128K |
| gpt-3.5-turbo-0613 | ChatGPT | Balanced | 4K |
| gpt-3.5-turbo-16k | ChatGPT | Balanced | 16K |
| gpt-4 | ChatGPT | Frontier | 8K |
Protocol support
Claude models are native to the Anthropic Messages API; GPT models are native to the OpenAI Responses API. Chat Completions works with every model, which is what most third-party tools use.
| Endpoint | Claude models | GPT models |
|---|---|---|
| POST /v1/messages | Yes | No |
| POST /v1/responses | No | Yes |
| POST /v1/chat/completions | Yes | Yes |
| GET /v1/models | Yes | Yes |
Choosing a tier
Most models in a tier bill at the same rate (the newest, Claude Fable 5.2, Mythos 5, Opus 5.5, Haiku 5.5 and GPT-6 Astra, at 1.5x), so choose by capability:
Frontier
The most capable models, for deep reasoning and long-horizon agent work.
e.g. claude-fable-5-2, claude-mythos-5, claude-opus-5.5, gpt-astra-6, gemini-argon-4, gemini-3-pro, gpt-5, deepseek-v4, deepseek-r1, grok-4, kimi-k2-thinking, kimi-k2, qwen-3-max, mistral-large-3, llama-4-maverick
Balanced
Strong everyday models that trade a little depth for speed and cost.
e.g. claude-sonnet-5.5, gemini-4-carbon, gemini-2.5-flash, gpt-5-mini, qwen-3-coder, codestral, llama-4-scout
Speed
Fast, lightweight models for high-volume and latency-sensitive work.
e.g. claude-haiku-5.5, gemini-2.5-flash-lite, mistral-small
A common pattern: plan and review with a Frontier model, do the bulk of the work with a Balanced one, and fan out sub-agents or classification on a Speed model. Claude and ChatGPT models can be mixed freely.
In the dashboard chat, smart routing can answer a simple follow-up with a lighter model, marked Auto; you can turn it off. See How routing works.