NovitaAI Provider Setup and Model Catalog for OpenClaw
Learn how to set up NovitaAI's OpenAI-compatible API with OpenClaw using the official plugin. Includes configuration steps, defaults, and a list of available models.
Read this when
- You want to run OpenClaw with NovitaAI models
- You need the Novita provider id, key, or endpoint
NovitaAI offers hosted AI infrastructure with an API that is compatible with OpenAI.
OpenClaw exposes NovitaAI through the official external
@openclaw/novita-provider plugin. Model references follow the
novita/deepseek/deepseek-v4-pro format.
Setup
Install the plugin, then restart the Gateway:
openclaw plugins install @openclaw/novita-provider
openclaw gateway restart
Generate an API key at novita.ai/settings/key-management, and afterwards execute:
openclaw onboard --auth-choice novita-api-key
Alternatively, configure:
export NOVITA_API_KEY="<your-novita-api-key>" # pragma: allowlist secret
Defaults
| Setting | Value |
|---|---|
| Plugin | @openclaw/novita-provider |
| Provider id | novita |
| Aliases | novita-ai, novitaai |
| Base URL | https://api.novita.ai/openai/v1 |
| Env var | NOVITA_API_KEY |
| Default model | novita/deepseek/deepseek-v4-pro |
Model catalog
novita/moonshotai/kimi-k3novita/moonshotai/kimi-k2.7-codenovita/minimax/minimax-m3novita/zai-org/glm-5.2novita/deepseek/deepseek-v4-pronovita/deepseek/deepseek-v4-flashnovita/qwen/qwen3.7-max
novita/minimax/minimax-m2.7 is still available as a legacy compatibility option, but it is not shown in model selection menus.
Treat this list as a baseline, not an up-to-date inventory. Your account, region, or Novita's present offerings might add, remove, or limit routes. Verify before locking in a default that persists:
openclaw models list --provider novita
When to choose Novita
- Access to hosted open-weight models through an OpenAI-compatible API.
- DeepSeek, Kimi, MiniMax, GLM, or Qwen-family routes under a single provider account.
- An additional hosted fallback option alongside DeepInfra, GMI, OpenRouter, or direct vendor APIs.
- Provider-side model hosting, removing the need to run LM Studio, Ollama, SGLang, or vLLM infrastructure yourself.
Pick a direct vendor provider when you require vendor-specific request parameters or support agreements. Pick a local provider when the model has to run on your own hardware or within your network boundary.
Troubleshooting
401/403: confirm the key on Novita's key management page and re-runopenclaw onboard --auth-choice novita-api-keyif the saved profile is outdated.- Unknown model errors: use the exact
novita/<route-id>returned byopenclaw models list --provider novita. - Slow or failed routes: switch to another Novita model route, or designate Novita as a fallback provider for workloads that can handle provider-specific variability.