gpuos

Integration · Coding assistant

Continue with your own GPUs

Use self-hosted coding models in VS Code and JetBrains through Continue.

Base URL
https://gpuos.si/v1
API key
gpuos_key_… from the dashboard
Model
a catalog id, e.g. qwen3-32b

Add gpuos models to Continue

Add models with the openai provider and your gpuos endpoint as apiBase in Continue's config.yaml.

~/.continue/config.yaml
models:
  - name: gpt-oss 20B (gpuos)
    provider: openai
    model: gpt-oss-20b
    apiBase: https://gpuos.si/v1
    apiKey: ${{ secrets.GPUOS_API_KEY }}
    roles: [chat, edit, apply]
  - name: Qwen3 8B (gpuos)
    provider: openai
    model: qwen3-8b
    apiBase: https://gpuos.si/v1
    apiKey: ${{ secrets.GPUOS_API_KEY }}
    roles: [autocomplete]

gpt-oss 20B is tuned for agentic coding and fits in about 14 GB; a small model like Qwen3 8B keeps autocomplete fast.

Official documentation: www.continue.dev

Questions

Does my code leave my machines?
Prompts go from your editor to the gpuos gateway in the EU, which forwards them to your own GPU node. gpuos stores token counts and latency for usage reports, not the code or the completions.
Can the whole team share the same models?
Yes. Create one key per developer, each with its own monthly token quota and rate limit.

Related

Use Continue with models on your GPUs

Free for one GPU. Connect a machine, deploy a model, then paste the base URL and your key.

Start freeRead the quickstart