OpenClaw + Nebius Token Factory

OpenClaw is a self-hosted AI assistant that connects chat apps (WhatsApp, Telegram, Discord, and more) to LLM providers. Because Nebius Token Factory serves open models through an OpenAI-compatible API, you can point OpenClaw at Token Factory and run your agents on models like Qwen3, Llama, and DeepSeek.

Why Token Factory?

OpenAI-compatible — no special SDK, OpenClaw's built-in OpenAI provider works as-is.
Open models — Qwen3, Kimi, DeepSeek, GLM and more, all behind one API key.
Cost control — usage-based pricing with no infra to manage.

Prerequisites

A running OpenClaw instance.
A Nebius Token Factory API key — see Getting Started for how to create one.

1. Get your API key

Grab a key from tokenfactory.nebius.com

2. Add Token Factory as a provider

openclaw --configure

Choose 'Configure Models'
choose 'Custom provider'
base url = https://api.tokenfactory.nebius.com/v1
API KEY = paste your nebius API key
Select OpenAI comptible
model id = moonshotai/Kimi-K2.6 (or any model id from Token Factory)
End point id = nebius-token-factory
model alias : leave blank

Model Selection

Model IDs use the org/model form. A few options:

Model	ID
Kimi-K2.6	`moonshotai/Kimi-K2.6`
DeepSeek-V4-Pro	`deepseek-ai/DeepSeek-V4-Pro`
GLM-5.1	`zai-org/GLM-5.1`

See the full catalog at tokenfactory.nebius.com or the Models guides in this cookbook.

That's it!

3. Restart OpenClaw

Be sure to restart the openclaw gateway for changes to take effect.

openclaw  gateway run
# or 
openclaw  gateway restart

Then launch openclaw as

# web based UI
openclaw  dashboard

# Terminal based ui
openclaw  tui

Troubleshooting

401 Unauthorized — check that NEBIUS_API_KEY is set and the provider config references it correctly.
404 / model not found — verify the model ID (org/model, case-sensitive) and that it's available on your account.
Wrong endpoint — the base_url must end in /v1; do not append /chat/completions yourself.