Integration · Agent
Use any model in OpenCode through one custom provider
OpenCode is an open-source terminal coding agent that supports custom providers in opencode.json. AutoJev writes an autojev provider for you, lists every model and route you select under it, and routes each request to the right upstream.
- One autojev provider in opencode.json(c)
- All selected models and routes listed
- Chat Completions in, any protocol out
- Original config restored on disconnect
Agents
CodexResponses
~/.codex/config.toml
Claude CodeMessages
~/.claude/settings.json
Gemini CLIGemini
~/.gemini/settings.json
OpenCodeChat
~/.config/opencode/opencode.json
What changes in OpenCode
Normally you add each provider to OpenCode by hand, with its own base URL, key and model list. With AutoJev, OpenCode gets exactly one provider — autojev — pointing at http://127.0.0.1:9527/v1 through the @ai-sdk/openai-compatible package. Every model and route you select on the Agents page becomes a model under that provider.
OpenCode talks Chat Completions to the gateway. AutoJev then forwards each request to the model’s real provider, converting to Responses or Anthropic Messages if that upstream needs it. Provider keys stay in AutoJev; OpenCode only holds a placeholder key that labels its requests in your logs.
What AutoJev configures
What OpenCode gains from the gateway
One place for keys
Provider keys live in AutoJev on your machine. OpenCode 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 opencode, 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 OpenCode’s config again.
Set up OpenCode with AutoJev
- 1
Install AutoJev
Download the desktop app and open it; the local gateway starts on port 9527.
- 2
Add providers and models
Add DeepSeek, OpenRouter, OpenAI, Ollama or any OpenAI-compatible or Anthropic-compatible endpoint with your own key, test it, and add models.
- 3
Optional: create routes
Routes like autojev/fast, autojev/quality or autojev/cheap let AutoJev choose the model per request with load-balancing or intelligent-selection strategies.
- 4
Connect OpenCode
On Agents, AutoJev detects the opencode executable. Select models and routes, pick the default, and click Connect.
- 5
Start OpenCode
Your selections appear under the AutoJev provider in OpenCode’s model list.
bash
The config AutoJev writes
Example with a route (autojev/fast) and a single model selected. OpenCode model IDs are provider-prefixed, so the default route becomes autojev/autojev/fast.
Switching models and routes
Pick any of the AutoJev models from OpenCode’s model list during a session. To add or remove options, or change the default, update the selection on the Agents page and reconnect — or use “Batch update configurations” to refresh every connected agent. Restart OpenCode to reload the list.
Connecting routes rather than individual models keeps OpenCode’s config stable: change the route in AutoJev and the next request uses the new rule.
Restoring opencode.json
AutoJev validates the file before writing and keeps a backup of the original. On disconnect, “Stop and restore configurations” or quit, it reverts what it wrote and keeps non-conflicting edits you made since. JSONC comments may be reformatted while connected.
Every agent on one page
OpenCode sits next to Codex, Claude Code and your other agents, each with its config path, selected models and connection status.
Agents
CodexResponses
~/.codex/config.toml
Claude CodeMessages
~/.claude/settings.json
Gemini CLIGemini
~/.gemini/settings.json
OpenCodeChat
~/.config/opencode/opencode.json
Frequently asked questions
How do I add a custom provider to OpenCode?
With AutoJev you don’t edit opencode.json by hand: connect OpenCode on the Agents page and AutoJev adds the autojev provider with your selected models. Every provider you add in AutoJev becomes reachable through it.
Can OpenCode use DeepSeek through AutoJev?
Yes. Add the DeepSeek preset and model in AutoJev, select it for OpenCode and connect. It shows up as autojev/<provider>/<model_id> in OpenCode.
Does AutoJev work with opencode.jsonc?
Yes. If opencode.jsonc exists, AutoJev edits it instead of opencode.json. OPENCODE_CONFIG also overrides the path.
Why is the model ID autojev/autojev/fast?
OpenCode prefixes model IDs with the provider name. The provider is autojev and the route ID is autojev/fast.
What about Kilo?
Kilo uses the same configuration layout. Add it as a custom agent and AutoJev recognizes the Kilo template automatically.
Keep exploring
Custom agents & BYOK clientsAdd any local agent executable, use templates for Kilo and more, or copy env, JSON and cURL connection parameters.FeaturesOne-click agent setupConnect Codex, Claude Code, Gemini CLI, OpenCode and more. AutoJev edits each native config, backs it up and restores it.FeaturesSmart model routingExpose routes like autojev/fast to every agent. Load-balance models or let Jev pick by cost, quality, speed or balance.Integrations
OpenRouterBring OpenRouter models to every agent. The same key can power Jev, the decision model for intelligent routing.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.
Free & open source · macOS · Windows · Linux