# Twigg model catalogue

Every active model you can name as `model` on `POST /api/v1/chats/{chat_id}/responses`. Prices are USD per million tokens and are what you pay: they include the platform margin of 5% over provider cost.

Reasoning tokens are a subset of output and are not billed separately. Compaction runs are billed as their own line at the same rates. Send the `name` column, never the provider's own model string.

| name | provider | context | max output | reasoning | tools | input | output | cache read | cache write |
|---|---|---|---|---|---|---|---|---|---|
| `claude-fable-5-1` | Anthropic | 1M | 128k | yes | yes | $10.5 | $52.5 | $0.2625 | $13.125 |
| `claude-opus-5` | Anthropic | 1M | 128k | yes | yes | $5.25 | $26.25 | $0.525 | $6.5625 |
| `claude-sonnet-5` | Anthropic | 1M | 128k | yes | yes | $2.1 | $10.5 | $0.21 | $2.625 |
| `gpt-5-6-luna` | OpenAI | 1M | 128k | yes | yes | $0.21 | $1.26 | $0.021 | $1.3125 |
| `gpt-5-6-sol` | OpenAI | 1M | 128k | yes | yes | $2.1 | $10.5 | $0.21 | — |
| `gpt-5-6-terra` | OpenAI | 1M | 128k | yes | yes | $2.1 | $12.6 | $0.21 | — |
| `gpt-6-astra` | OpenAI | 1.1M | 32k | yes | yes | $10.5 | $52.5 | $1.05 | $13.125 |
| `gemini-3.1-flash-lite` | Google | 1.0M | 65k | yes | yes | $0.2625 | $1.575 | $0.02625 | — |
| `gemini-3.1-pro-preview` | Google | 1.0M | 65k | yes | yes | $2.1 | $12.6 | $0.21 | — |
| `gemini-3.8-flash` | Google | 1.0M | 65k | yes | yes | $1.575 | $7.875 | $0.1575 | — |
| `grok-4-3` | xAI | 1M | 32k | yes | yes | $1.3125 | $2.625 | $0.21 | — |
| `grok-4-5` | xAI | 500k | 32k | yes | yes | $2.1 | $6.3 | $0.315 | — |
| `grok-4-6` | xAI | 500k | 128k | yes | yes | $2.1 | $6.3 | $0.525 | — |
| `fireworks/deepseek-v4.1-flash` | Fireworks AI | 1M | 32k | yes | yes | $0.231 | $0.693 | $0.00735 | — |
| `fireworks/deepseek-v4-flash-0731` | Fireworks AI | 1.0M | 32k | yes | yes | $0.231 | $0.693 | $0.00735 | — |
| `fireworks/deepseek-v4-flash-vision-exp` | Fireworks AI | 1.0M | 32k | yes | yes | $0.231 | $0.693 | $0.00735 | — |
| `fireworks/deepseek-v4-pro-0813` | Fireworks AI | 1.0M | 32k | yes | yes | $1.386 | $4.158 | $0.0462 | — |
| `fireworks/glm-5.2` | Fireworks AI | 1.0M | 32k | yes | yes | $1.47 | $4.62 | $0.147 | — |
| `fireworks/glm-5.3` | Fireworks AI | 1M | 32k | yes | yes | $1.47 | $4.62 | $0.273 | — |
| `fireworks/glm-5.3-flash` | Fireworks AI | 1.0M | 32k | yes | yes | $0.1575 | $0.525 | $0.0315 | — |
| `fireworks/gpt-oss-120b` | Fireworks AI | 128k | 32k | yes | yes | $0.1575 | $0.63 | $0.01575 | — |
| `fireworks/inkling` | Fireworks AI | 1.0M | 32k | yes | yes | $1.05 | $4.2525 | $0.1785 | — |
| `fireworks/kimi-k2.6` | Fireworks AI | 256k | 32k | yes | yes | $0.9975 | $4.2 | $0.168 | — |
| `fireworks/kimi-k2.7-code` | Fireworks AI | 262k | 32k | yes | yes | $0.9975 | $4.2 | $0.1995 | — |
| `fireworks/kimi-k3` | Fireworks AI | 1.0M | 32k | yes | yes | $3.15 | $15.75 | $0.315 | — |
| `fireworks/minimax-m3` | Fireworks AI | 512k | 32k | yes | yes | $0.315 | $1.26 | $0.063 | — |
| `fireworks/muse-glimmer-30b` | Fireworks AI | 131k | 32k | yes | yes | $0.3675 | $1.575 | $0.042 | — |
| `fireworks/nemotron-3-ultra-nvfp4` | Fireworks AI | 262k | 32k | yes | yes | $0.63 | $2.52 | $0.126 | — |
| `fireworks/nemotron-lightning-3.5-30b-a3b` | Fireworks AI | 262k | 32k | yes | yes | $0.0525 | $0.21 | $0.0105 | — |
| `fireworks/qwen3.8-max` | Fireworks AI | 256k | 32k | yes | yes | $2.1 | $6.3 | $0.2625 | — |
| `openrouter/claude-haiku-4-5` | OpenRouter | 200k | 64k | yes | yes | $1.05 | $5.25 | $0.105 | $1.3125 |
| `openrouter/claude-sonnet-5` | OpenRouter | 1M | 128k | yes | yes | $2.1 | $10.5 | $0.21 | $2.625 |
| `openrouter/gpt-5-6-sol` | OpenRouter | 1M | 128k | yes | yes | $2.1 | $10.5 | $0.21 | — |
| `openrouter/grok-4-6` | OpenRouter | 500k | 128k | yes | yes | $2.1 | $6.3 | $0.525 | — |
| `openrouter/hy4-preview` | OpenRouter | 1.0M | 32k | yes | yes | $0.8757 | $2.62605 | $0.0441 | — |
| `openrouter/mimo-v2.5` | OpenRouter | 1.0M | 32k | yes | yes | $0.147 | $0.294 | $0.00294 | — |
| `openrouter/mimo-v2.5-pro` | OpenRouter | 1.0M | 32k | yes | yes | $0.45675 | $0.9135 | $0.00378 | — |
| `openrouter/minimax-m3` | OpenRouter | 524k | 32k | yes | yes | $0.315 | $1.26 | $0.063 | — |
| `openrouter/mistral-large-2512` | OpenRouter | 262k | 32k | no | yes | $0.525 | $1.575 | $0.0525 | — |
| `openrouter/nemotron-3-ultra-550b-a55b` | OpenRouter | 202k | 32k | yes | yes | $0.63 | $2.52 | $0.126 | — |
| `openrouter/qwen3.8-flash` | OpenRouter | 1M | 32k | yes | yes | $0.1575 | $0.4935 | $0.0168 | $0.21 |
| `openrouter/qwen3-coder-next` | OpenRouter | 262k | 32k | no | yes | $0.126 | $0.84 | $0.0735 | — |

## Attachments

Configured models reject unsupported new attachments with HTTP 422 before persistence. Retention runs first on history; incompatible retained attachments become text markers with warnings while stored bytes remain intact. The total HTTP request body limit also applies (HTTP 413), including base64 overhead.

- `claude-fable-5-1`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (7500000 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (7500000 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `claude-opus-5`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (7500000 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (7500000 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `claude-sonnet-5`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (7500000 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (7500000 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `gpt-5-6-luna`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `gpt-5-6-sol`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `gpt-5-6-terra`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `gpt-6-astra`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); tool result: image/png, image/jpeg, image/gif, image/webp, application/pdf, text/plain (20971520 bytes/attachment); assistant: unsupported; 20 attachments/part.
- `gemini-3.1-flash-lite`: user: image/png, image/jpeg, image/webp, application/pdf, text/plain (7000000 bytes/attachment, 3000/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `gemini-3.1-pro-preview`: user: image/png, image/jpeg, image/webp, application/pdf, text/plain (7000000 bytes/attachment, 3000/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `gemini-3.8-flash`: user: image/png, image/jpeg, image/webp, application/pdf, text/plain (7000000 bytes/attachment, 3000/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `grok-4-3`: user: image/png, image/jpeg (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `grok-4-5`: user: image/png, image/jpeg (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `grok-4-6`: user: image/png, image/jpeg (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/deepseek-v4.1-flash`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/deepseek-v4-flash-0731`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/deepseek-v4-flash-vision-exp`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/deepseek-v4-pro-0813`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/glm-5.2`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/glm-5.3`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/glm-5.3-flash`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/gpt-oss-120b`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/inkling`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/kimi-k2.6`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/kimi-k2.7-code`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/kimi-k3`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/minimax-m3`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/muse-glimmer-30b`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/nemotron-3-ultra-nvfp4`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/nemotron-lightning-3.5-30b-a3b`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `fireworks/qwen3.8-max`: user: image/png, image/jpeg, image/gif (20971520 bytes/attachment, 30/request including history); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/claude-haiku-4-5`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf (3750000 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/claude-sonnet-5`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf (3750000 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/gpt-5-6-sol`: user: image/png, image/jpeg, image/gif, image/webp, application/pdf (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/grok-4-6`: user: image/png, image/jpeg, application/pdf (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/hy4-preview`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/mimo-v2.5`: user: image/png, image/jpeg, image/gif, image/webp (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/mimo-v2.5-pro`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/minimax-m3`: user: image/png, image/jpeg, image/gif, image/webp (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/mistral-large-2512`: user: image/png, image/jpeg, image/gif, image/webp (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/nemotron-3-ultra-550b-a55b`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/qwen3.8-flash`: user: image/png, image/jpeg, image/gif, image/webp (20971520 bytes/attachment); tool result: unsupported; assistant: unsupported; 20 attachments/part.
- `openrouter/qwen3-coder-next`: user: unsupported; tool result: unsupported; assistant: unsupported; 20 attachments/part.

## Pricing conditions: `gpt-6-astra`

- The whole request reprices when context tokens above 272k: input $21, output $78.75, cache read $2.1, cache write $26.25.

## Pricing conditions: `gemini-3.1-pro-preview`

- The whole request reprices when context tokens above 200k: input $4.2, output $18.9, cache read $0.42.

## Pricing conditions: `grok-4-3`

- The whole request reprices when context tokens above 200k: input $2.625, output $5.25, cache read $0.42.

## Pricing conditions: `grok-4-5`

- The whole request reprices when context tokens above 200k: input $4.2, output $12.6, cache read $0.63.

JSON: `GET https://api.twigg.ai/v1/catalogue`. Per model: `GET https://api.twigg.ai/v1/catalogue/{name}`.