AutoJev
Download

Integration · Agent

Use any model in OpenClaw through AutoJev

OpenClaw reads its model providers from ~/.openclaw/openclaw.json. AutoJev registers itself there as one provider, lists the models and routes you select, sets the primary model, and routes each request to the provider behind it.

  • Adds models.providers.autojev
  • Sets agents.defaults.model.primary
  • Chat Completions in, any protocol out
  • Restores openclaw.json on disconnect
Download AutoJevStar on GitHub

Agents

+ Agent

CodexResponses

~/.codex/config.toml

Connected

Claude CodeMessages

~/.claude/settings.json

Connected

Gemini CLIGemini

~/.gemini/settings.json

Connected

OpenCodeChat

~/.config/opencode/opencode.json

Connect

How AutoJev connects OpenClaw

OpenClaw separates model providers from the agent defaults that pick a model. AutoJev writes both: a provider named autojev under models.providers, pointing at http://127.0.0.1:9527/v1 with the openai-completions API, and the default agent model under agents.defaults.

Each model or route you select becomes an entry in the provider’s models list (with a 128k context window and 8,192 max tokens set as defaults) and is also allowed under agents.defaults.models, so you can switch to it from OpenClaw. From there, AutoJev routes to the real upstream and converts protocols as needed.

What AutoJev configures

ItemValue
Config file~/.openclaw/openclaw.json (OPENCLAW_CONFIG_PATH / OPENCLAW_STATE_DIR respected)
Providermodels.providers.autojev, api = openai-completions
Base URLhttp://127.0.0.1:9527/v1
Keyautojev-local-openclaw (placeholder; labels requests as openclaw)
Primary modelagents.defaults.model.primary = autojev/<model or route id>
Allowed modelsagents.defaults.models gets one autojev/… entry per selection

What OpenClaw gains from the gateway

One place for keys

Provider keys live in AutoJev on your machine. OpenClaw only holds a placeholder key, and requests go directly to your providers with no markup.

Failover before output

On connection errors or HTTP 401/402/403/408/429/5xx before output starts, a route retries another candidate (up to 4 attempts by default) and respects Retry-After.

Per-agent logs and cost

Every request is labeled openclaw, with model, route, input/output/cache tokens, latency and estimated cost. Prompts and responses are not stored.

Routing without edits

Connect a route once. Switch its strategy between load balancing and intelligent selection in AutoJev without touching OpenClaw’s config again.

Set up OpenClaw with AutoJev

  1. 1

    Install AutoJev

    Download the desktop app and open it; the gateway listens on 127.0.0.1:9527.

  2. 2

    Add providers and models

    Add a preset (OpenAI, DeepSeek, OpenRouter, Ollama) or any OpenAI-compatible or Anthropic-compatible endpoint with your key, then add models.

  3. 3

    Optional: create a route

    Use a route like autojev/cheap or autojev/quality when you want AutoJev to choose or fail over between models.

  4. 4

    Connect OpenClaw

    On Agents, AutoJev detects the openclaw executable. Select models and routes, choose the default and click Connect.

  5. 5

    Restart OpenClaw

    Restart OpenClaw so it loads the new provider and primary model.

    bash
    openclaw

The config AutoJev writes

Example with the route autojev/fast selected as default. OpenClaw model references are provider-prefixed, hence autojev/autojev/fast.

~/.openclaw/openclaw.json
{ "models": { "providers": { "autojev": { "baseUrl": "http://127.0.0.1:9527/v1", "apiKey": "autojev-local-openclaw", "api": "openai-completions", "models": [ { "id": "autojev/fast", "name": "Fast", "contextWindow": 128000, "maxTokens": 8192 } ] } } }, "agents": { "defaults": { "model": { "primary": "autojev/autojev/fast" }, "models": { "autojev/autojev/fast": {} } } } }

Switching models and routes

To change the primary model or the list of available ones, update the selection on the Agents page and reconnect, or use “Batch update configurations”. AutoJev rewrites only its own autojev/… entries and leaves other providers and allowed models in place.

Connecting a route keeps openclaw.json stable — tune the route in AutoJev and OpenClaw keeps calling the same ID.

Restoring openclaw.json

AutoJev validates the file before writing and backs up the original. Disconnecting, “Stop and restore configurations” or quitting reverts AutoJev’s keys while keeping non-conflicting edits you made later. JSON5 comments may be reformatted while connected.

Frequently asked questions

How do I add a custom model provider to OpenClaw?

Connect OpenClaw on AutoJev’s Agents page. AutoJev adds an autojev provider to openclaw.json, and every provider you configure in AutoJev becomes reachable through it.

Can OpenClaw use DeepSeek through AutoJev?

Yes. Add DeepSeek in AutoJev, select the model for OpenClaw and connect. It appears as autojev/<provider>/<model_id>.

Which API does OpenClaw use with AutoJev?

OpenClaw is connected with api = openai-completions (Chat Completions). AutoJev converts to Responses or Anthropic Messages when the upstream model needs it.

Will AutoJev remove my other OpenClaw providers?

No. It only adds or updates the autojev provider and its own autojev/… model entries.

Where do the 128000 and 8192 values come from?

They are the context window and max tokens AutoJev writes for each entry by default; the actual limits are those of the upstream model.

Keep exploring

FeaturesOne-click agent setupConnect Codex, Claude Code, Gemini CLI, OpenCode and more. AutoJev edits each native config, backs it up and restores it.FeaturesProtocol conversionConvert between Chat Completions, Responses and Anthropic Messages in all six directions, with streaming and tool calls.IntegrationsCustom agents & BYOK clientsAdd any local agent executable, use templates for Kilo and more, or copy env, JSON and cURL connection parameters.IntegrationsDeepSeekRun DeepSeek in Codex and Claude Code. AutoJev converts Responses and Messages to Chat Completions on localhost.Use casesFor multi-agent developersCodex, Claude Code, Gemini CLI, OpenCode and more on one local gateway, with per-agent models and one request log.

Updated 2026-10-14

Give every agent the right model.

Install AutoJev, add the providers you already pay for, and connect your agents in one click. Everything stays on your machine.

Download AutoJevStar on GitHub

Free & open source · macOS · Windows · Linux