OpenClaw

Arbytra connects to OpenClaw as a custom model provider. You can use multiple models through a single API key.

Prerequisites

OpenClaw v2026.1.30+
Node.js 22.16+
An Arbytra API key

Install OpenClaw

Install via npm:

npm install -g openclaw@latest

Or use the installer script:

curl -fsSL https://openclaw.ai/install.sh | bash

If you used the installer script, open a new terminal. Then confirm:

openclaw --version

You need version v2026.1.30 or later. Start the gateway daemon:

openclaw onboard --install-daemon

The daemon runs on port 18789 and must be running for OpenClaw to process requests.

Set your API key

Export your API key:

export ARBYTRA_API_KEY="ak_live_..."

Your key starts with ak_live_ (production) or ak_test_ (testing). To persist across terminal sessions, add this line to ~/.zshrc or ~/.bashrc. Reload your shell after saving:

source ~/.zshrc   # or: source ~/.bashrc

Or open a new terminal. If you installed the daemon via openclaw onboard --install-daemon, also register the key with the daemon:

openclaw config set env.vars.ARBYTRA_API_KEY "$ARBYTRA_API_KEY"

The daemon runs as a LaunchAgent and doesn’t inherit your shell’s environment variables. Without this step, the daemon can’t resolve ${ARBYTRA_API_KEY} from the config file.

Configure the provider

Create or update ~/.openclaw/openclaw.json with the Arbytra provider:

{
  models: {
    mode: "merge",
    providers: {
      "arbytra": {
        baseUrl: "https://api.arbytra.com/v1",
        apiKey: "${ARBYTRA_API_KEY}",
        api: "openai-completions",
        agentRuntime: { id: "pi" },
        models: [
          { id: "claude-sonnet-4-6", name: "Claude Sonnet 4.6", contextWindow: 1000000, maxTokens: 64000 },
          { id: "claude-opus-4-6", name: "Claude Opus 4.6", contextWindow: 1000000, maxTokens: 128000 },
          { id: "claude-opus-4-7", name: "Claude Opus 4.7", contextWindow: 1000000, maxTokens: 128000 },
          { id: "deepseek-v4-flash", name: "DeepSeek V4 Flash", contextWindow: 1000000, maxTokens: 8192 },
          { id: "deepseek-v4-pro", name: "DeepSeek V4 Pro", contextWindow: 1000000, maxTokens: 8192 },
          { id: "gemini-2.5-flash", name: "Gemini 2.5 Flash", contextWindow: 1000000, maxTokens: 65536 },
          { id: "gemini-2.5-pro", name: "Gemini 2.5 Pro", contextWindow: 1000000, maxTokens: 65536 },
          { id: "gemini-3.1-pro-preview", name: "Gemini 3.1 Pro Preview", contextWindow: 1000000, maxTokens: 65536 },
          { id: "glm-5.1", name: "GLM-5.1", contextWindow: 200000, maxTokens: 8192 },
          { id: "grok-4.3", name: "Grok 4.3", contextWindow: 1000000, maxTokens: 8192 },
          { id: "kimi-k2.5", name: "Kimi K2.5", contextWindow: 262144, maxTokens: 8192 },
          { id: "kimi-k2.6", name: "Kimi K2.6", contextWindow: 262144, maxTokens: 8192 },
          { id: "minimax-m2-7", name: "MiniMax M2.7", contextWindow: 205000, maxTokens: 8192 },
          { id: "minimax-m2-7-highspeed", name: "MiniMax M2.7 Highspeed", contextWindow: 205000, maxTokens: 8192 },
          { id: "qwen-3.6-plus", name: "Qwen 3.6 Plus", contextWindow: 1000000, maxTokens: 8192 }
        ]
      }
    }
  },
  agents: {
    defaults: {
      model: {
        primary: "arbytra/claude-sonnet-4-6"
      },
      models: {
        "arbytra/claude-sonnet-4-6": {},
        "arbytra/claude-opus-4-6": {},
        "arbytra/claude-opus-4-7": {},
        "arbytra/deepseek-v4-flash": {},
        "arbytra/deepseek-v4-pro": {},
        "arbytra/gemini-2.5-flash": {},
        "arbytra/gemini-2.5-pro": {},
        "arbytra/gemini-3.1-pro-preview": {},
        "arbytra/glm-5.1": {},
        "arbytra/grok-4.3": {},
        "arbytra/kimi-k2.5": {},
        "arbytra/kimi-k2.6": {},
        "arbytra/minimax-m2-7": {},
        "arbytra/minimax-m2-7-highspeed": {},
        "arbytra/qwen-3.6-plus": {}
      }
    }
  }
}

The models array uses bare IDs (claude-sonnet-4-6) matching what Arbytra’s API expects. The agents.defaults.models object uses prefixed IDs (arbytra/claude-sonnet-4-6). OpenClaw doesn’t route to a model unless it appears in both places. models.mode: "merge" keeps OpenClaw’s bundled providers alongside Arbytra. If you already have a config from openclaw onboard, merge the arbytra provider into your existing models.providers object and add the arbytra/ entries to your existing agents.defaults.models. ${ARBYTRA_API_KEY} reads your environment variable. Don’t paste your key into the config. agentRuntime: { id: "pi" } selects the embedded agent harness. Without it, OpenClaw defaults to codex which requires a separate binary.

Verify

Test that the connection works:

openclaw agent --message "Say exactly: setup-ok"

Use different models

Switch the active model from the CLI:

openclaw models set arbytra/deepseek-v4-flash

To switch models during an interactive session, type /model arbytra/<id>. Models available through Arbytra:

Model	Author	Context
`claude-sonnet-4-6`	Anthropic	1M
`claude-opus-4-6`	Anthropic	1M
`claude-opus-4-7`	Anthropic	1M
`deepseek-v4-flash`	DeepSeek	1M
`deepseek-v4-pro`	DeepSeek	1M
`gemini-2.5-flash`	Google	1M
`gemini-2.5-pro`	Google	1M
`gemini-3.1-pro-preview`	Google	1M
`glm-5.1`	Z.AI	200K
`grok-4.3`	xAI	1M
`kimi-k2.5`	Moonshot	262K
`kimi-k2.6`	Moonshot	262K
`minimax-m2-7`	MiniMax	205K
`minimax-m2-7-highspeed`	MiniMax	205K
`qwen-3.6-plus`	Alibaba	1M

To use a model not in this list, add its bare ID to models.providers.arbytra.models and its prefixed ID to agents.defaults.models in your config. To list all Arbytra models:

curl -s -H "Authorization: Bearer $ARBYTRA_API_KEY" \
  https://api.arbytra.com/v1/models | jq '.data[].id'

Run a health check

OpenClaw includes a diagnostic command:

openclaw doctor

This checks daemon connectivity, provider configuration, and model availability. Run openclaw doctor --fix to auto-resolve common problems.

Control routing

Configure routing in the Arbytra dashboard. See routing options for details.

Uninstall

Remove the arbytra provider block from ~/.openclaw/openclaw.json and delete any arbytra/ entries from agents.defaults.models. Remove ARBYTRA_API_KEY from your shell profile (~/.zshrc or ~/.bashrc).

Troubleshoot

Symptom	Cause	Fix
`Error: API key is invalid.`	`ARBYTRA_API_KEY` not set or wrong	Run `echo $ARBYTRA_API_KEY`. If empty, export it. If set, verify at `https://api.arbytra.com/v1/me`
`Error: Model not found: arbytra/<id>`	Model not in `agents.defaults.models` or provider’s `models` array	Add the model to both places in `~/.openclaw/openclaw.json`
Connection refused on port 18789	Daemon isn’t running	Run `openclaw gateway status`. If stopped, run `openclaw onboard --install-daemon`
`openclaw: command not found`	npm global bin not in PATH	Run `export PATH="$(npm prefix -g)/bin:$PATH"`
Config parse error on startup	Invalid JSON5 in `openclaw.json`	Run `openclaw config schema` to validate. Check for mismatched braces
Empty response from agent	Model missing `contextWindow` in provider config	Add `contextWindow` and `maxTokens` to the model definition
Duplicated response content (`setup-oksetup-ok`)	OpenClaw `pi` harness concatenation bug	Known upstream issue. Responses are correct but repeated. Doesn’t affect all models equally. No user-side fix available.
API key valid in shell but agent requests fail	Daemon doesn’t inherit shell env vars	Run `openclaw config set env.vars.ARBYTRA_API_KEY "$ARBYTRA_API_KEY"`

Get Started

Guides

Response API

Frameworks

Integrations

Specifications

Platform

Changelog

Prerequisites

Install OpenClaw

Set your API key

Configure the provider

Verify

Use different models

Control routing

Uninstall

Troubleshoot

​Prerequisites

​Install OpenClaw

​Set your API key

​Configure the provider

​Verify

​Use different models

​Control routing

​Uninstall

​Troubleshoot

Prerequisites

Install OpenClaw

Set your API key

Configure the provider

Verify

Use different models

Control routing

Uninstall

Troubleshoot