Models & licences
Frontier models
Proprietary models. They run under the terms their provider sets, not under an open licence, and a team brings its existing plan for them onto the platform.
| Provider | Models | Origin | Licence |
|---|---|---|---|
| OpenAI | GPT-6 Astra, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna | USA | Provider terms |
| Anthropic | Fable 5.1, Opus 5, Sonnet 5, Haiku 4.5 | USA | Provider terms |
| Google DeepMind | Gemini 3.8 Pro, Gemini 3.8 Flash | USA | Provider terms |
| xAI | Grok 4.7, Grok 4.6 | USA | Provider terms |
| Cursor | Composer | USA | Provider terms |
OpenAI
Anthropic
Google DeepMind
xAI
CursorOpen-weight models
Published weights, each under its own licence. The table below this one says which of them can receive a request today and at what price.
| Provider | Models | Origin | Licence |
|---|---|---|---|
| DeepSeek | DeepSeek-V4-Pro, DeepSeek-V4.1-Flash, DeepSeek-R1 | China | MIT |
| Z.ai (Zhipu AI) | GLM 5.2, GLM 4.6 | China | MIT |
| Alibaba (Qwen) | Qwen3.6, Qwen3.5 397B-A17B, Qwen3-Coder-480B | China | Apache-2.0 |
| Moonshot AI | Kimi K2.6, Kimi K2 Thinking | China | Modified MIT |
| Mistral AI | Mistral Large 4, Mistral Medium 3.5, Devstral 2, Mistral Nemo | France | Mixed , see model |
| Meta | Llama 4 Maverick, Llama 4 Scout, Llama 3.3 70B | USA | Llama Community |
| Google DeepMind | Gemma 4, Gemma 3 27B | USA | Gemma Terms |
| NVIDIA | Nemotron 3 Ultra, Nemotron 3 Super | USA | NVIDIA Nemotron Open Model |
| OpenAI | gpt-oss 120b, gpt-oss 20b | USA | Apache-2.0 |
DeepSeek
Z.ai (Zhipu AI)
Alibaba (Qwen)
Moonshot AI
Mistral AI
Meta
Google DeepMind
NVIDIAModel versions checked on 7 October 2026. The field moves weekly; check the provider before relying on a version named here.
Shipped registry
| Model | Provider id | Vendor | Serving & price | Status | Licence |
|---|---|---|---|---|---|
| GLM 5.2 | glm-5.2 | Z.ai (Zhipu AI) | Scaleway Generative APIs — ServerlessPay per use · €1.80 input / €5.50 output per 1M tokens | Selectable | MIT |
| Qwen3.5 397B-A17B | qwen3.5-397b-a17b | Alibaba (Qwen) | Scaleway Generative APIs — ServerlessPay per use · €0.60 input / €3.60 output per 1M tokens | Selectable | Apache-2.0 |
| Qwen3.6 35B-A3B | qwen3.6-35b-a3b | Alibaba (Qwen) | Scaleway Generative APIs — ServerlessPay per use · €0.25 input / €1.50 output per 1M tokens | Selectable | Apache-2.0 |
| Qwen3 235B Instruct | qwen3-235b-a22b-instruct-2507 | Alibaba (Qwen) | Scaleway Generative APIs — ServerlessPay per use · €0.75 input / €2.25 output per 1M tokens | Selectable | Apache-2.0 |
| Mistral Medium 3.5 | mistral-medium-3.5-128b | Mistral AI | Scaleway Generative APIs — ServerlessPay per use · €1.50 input / €7.50 output per 1M tokens | Selectable | Modified MIT |
| gpt-oss 120b | gpt-oss-120b | OpenAI | Scaleway Generative APIs — ServerlessPay per use · €0.15 input / €0.60 output per 1M tokens | Selectable | Apache-2.0 |
| Llama 3.3 70B Instruct | llama-3.3-70b-instruct | Meta | Scaleway Generative APIs — ServerlessPay per use · €0.90 input / €0.90 output per 1M tokens | Selectable | Llama Community |
| Gemma 4 26B-A4B | gemma-4-26b-a4b-it | Scaleway Generative APIs — ServerlessPay per use · €0.25 input / €0.50 output per 1M tokens | Selectable | Apache-2.0 |
Models chosen by the server
Two jobs never go through the model switcher, so they are listed separately rather than left implicit.
- Chat titles Disabled pending attestation
- Public landing chat Disabled pending attestation