Product / Supported models
One platform. Every model.
Nordbase is model-agnostic: each agent picks its model, you swap freely, and tasks route to the model that earns its cost. This is the catalog you can run today.
OpenAI
The GPT line — the broad default for general work.
| Model | Model ID |
|---|---|
| GPT-6 Luna | gpt-6-luna |
| GPT-6 Sol | gpt-6-sol |
| GPT-5.6 Sol | gpt-5.6-sol |
| GPT-5.6 Terra | gpt-5.6-terra |
| GPT-5.6 Luna | gpt-5.6-luna |
| GPT-5.4 Mini | gpt-5.4-mini |
Anthropic
The Claude line — careful judgment over long, tool-heavy runs.
| Model | Model ID |
|---|---|
| Claude Fable 5 | claude-fable-5 |
| Claude Opus 5.5 | claude-opus-5-5 |
| Claude Opus 5 | claude-opus-5 |
| Claude Sonnet 5 | claude-sonnet-5 |
| Claude Opus 4.8 | claude-opus-4-8 |
| Claude Sonnet 4.6 | claude-sonnet-4-6 |
| Claude Haiku 4.5 | claude-haiku-4-5 |
The Gemini line — the Flash tiers carry high-volume work at speed.
| Model | Model ID |
|---|---|
| Gemini 3.8 Flash | gemini-3.8-flash |
| Gemini 3.7 Flash | gemini-3.7-flash |
| Gemini 3.1 Pro (preview) | gemini-3.1-pro-preview |
| Gemini 3.6 Flash | gemini-3.6-flash |
| Gemini 3.5 Flash | gemini-3.5-flash |
Mistral
Built in Europe — the natural pick when EU sourcing matters as much as EU hosting.
| Model | Model ID |
|---|---|
| Mistral Medium 3.5 | mistral-medium-3-5 |
| Mistral Small (2603) | mistral-small-2603 |
| Mistral Large (2512) | mistral-large-2512 |
| Ministral 3B (2512) | ministral-3b-2512 |
Moonshot AI
The Kimi line — strong capability at a markedly lower price.
| Model | Model ID |
|---|---|
| Kimi K3 | kimi-k3 |
| Kimi K2.7 Code | kimi-k2.7-code |
| Kimi K2.7 Code Highspeed | kimi-k2.7-code-highspeed |
| Kimi K2.6 | kimi-k2.6 |
Cerebras
Open weights on Cerebras's own hardware — built for latency, for agents where someone is waiting on the answer.
| Model | Model ID |
|---|---|
| GPT-OSS 120B | gpt-oss-120b |
| Qwen 3.8 27B | qwen-3.8-27b |
Infercom
Open weights on Infercom's own hardware. The first three are served from Munich; the rest run in their US and Japan regions.
| Model | Model ID | Region |
|---|---|---|
| MiniMax M2.7 | MiniMax-M2.7 | EU |
| GPT-OSS 120B | gpt-oss-120b | EU |
| Gemma 4 31B | gemma-4-31B-it | EU |
| DeepSeek V3.1 | DeepSeek-V3.1 | US / JP |
| DeepSeek V3.2 | DeepSeek-V3.2 | US / JP |
| Llama 3.3 70B | Meta-Llama-3.3-70B-Instruct | US / JP |
Self-hosted open models
Open models we run for you on dedicated, single-tenant endpoints in our EU cloud — no shared deployment, and your data never leaves Europe.
| Model | Model ID | Status |
|---|---|---|
| Qwen 3.5 32B | qwen3.5-32b | coming soon |
| Gemma 4 31B | gemma-4-31b | coming soon |
| Qwen 3.5 4B | qwen3.5-4b | available |
| Mistral Small 4 | mistral-small-4 | coming soon |
Don't see your model?
The catalog grows with our customers. If your company has standardised on a model we don't list yet, tell us — expanding the catalog is our side of the work.
Pick a model. Change your mind later.
Nothing on Nordbase marries you to a vendor — swap the model under an agent and keep everything else.