A local gateway and menu bar app that lets Claude Code, Codex, Gemini CLI and other coding agents use any model from any provider or subscription.
jasonxu114514/opencode2api
A Go gateway that exposes OpenCode Zen and Zen Go models through Chat Completions, Responses and Anthropic Messages APIs, with key and proxy pooling and a WebUI.
- Stars
- 1,099
- Forks
- 192
- Open issues
- 25
- Last commit
- 15 hours ago
What is it?
opencode2api is a self-hosted gateway for OpenCode Zen and Zen Go. It accepts requests in Chat Completions, Responses, or Anthropic Messages format and translates them into each upstream's native protocol, so clients built for one API style can reach models served under another. The project ships as a single Go binary with the WebUI embedded, and needs no Node.js runtime or database.
Upstream Zen and Go keys are managed in separate pools, with configurable tier preference, retries, session affinity, and direct, HTTP, HTTPS, SOCKS5, or SOCKS5H proxy routing. Model discovery refreshes from OpenCode's catalog, and a separate management port offers configuration editing, a Playground, diagnostics, token statistics, live logs, and a Go subscription quota dashboard.
The gateway supports text, images, reasoning, function tools, and tool results where the target protocol can represent them. It does not implement embeddings, file uploads, image generation, or conversation storage.
Key features
- 01Converts between Chat Completions, Responses and Anthropic Messages protocols, with JSON and SSE output.
- 02Separate Zen and Go key pools with tier preference, retries and session affinity.
- 03Routes through direct, HTTP, HTTPS, SOCKS5 or SOCKS5H proxies, including proxy files.
- 04Includes a WebUI on a separate port for configuration, Playground, diagnostics and live logs.
- 05Hot-reloads configuration, validating changes before new requests use them.
- 06Creates and manages local API keys for clients.
Use cases
- Pointing an OpenAI- or Anthropic-compatible coding tool at OpenCode Zen or Zen Go models.
- Sharing a pool of upstream keys and proxies across several local clients.
- Monitoring token usage and Go subscription quota from a single dashboard.
Topics
FAQ
- What is opencode2api?
- A Go gateway that exposes OpenCode Zen and Zen Go models through Chat Completions, Responses and Anthropic Messages APIs, with key and proxy pooling and a WebUI. opencode2api is a self-hosted gateway for OpenCode Zen and Zen Go. It accepts requests in Chat Completions, Responses, or Anthropic Messages format and translates them into each upstream's native protocol, so clients built for one API style can reach models served under another. The project ships as a single Go binary with the WebUI embedded, and needs no Node.js runtime or database.
- Is opencode2api free to use? What license does it have?
- opencode2api is open source under the GPL-3.0 license. Check the license terms before using it in commercial products.
- Is opencode2api still maintained?
- Actively maintained. The last commit to opencode2api was on Oct 19, 2026.
- What can I use opencode2api for?
- Pointing an OpenAI- or Anthropic-compatible coding tool at OpenCode Zen or Zen Go models. Sharing a pool of upstream keys and proxies across several local clients. Monitoring token usage and Go subscription quota from a single dashboard.
- Where can I get opencode2api?
- The source code and installation instructions are on GitHub: https://github.com/jasonxu114514/opencode2api
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Self-hosted, OpenAI-compatible LLM gateway that routes one endpoint to 40+ providers with virtual keys, budgets, and usage tracking.
Self-hostable MIT AI gateway that exposes one OpenAI-compatible endpoint to hundreds of model providers, with quota-aware auto-fallback and token compression.
A local model gateway that lets Claude Code, Codex, OpenCode and other coding agents share one endpoint and route requests across multiple model providers.
Open-source AI gateway giving one OpenAI-compatible API across 23+ LLM providers, with failover, load balancing, semantic caching and virtual keys.
A Go command-line tool and web UI for downloading models and datasets from the HuggingFace Hub with parallel connections and resume support.