GH Repository · tbphp
gpt-load
Self-hosted AI gateway for multi-channel, multi-credential setups — API keys and subscription accounts, scheduling, failover, request logs and usage. 自托管 AI 网关：多渠道多凭据统一接入，含密钥与订阅账号、调度容错、日志与用量。
- stars
- 6,809
- 30-day movement
- +14548/day
- Related entries
- 60
- Connections
- 1
makedockerotherself-hostedapi-gatewaygoopenaicodexgolanggeminiginanthropicclaudeai-gatewayclaude-codellmGollm-gatewayload-balancer
gpt-load is a self-hosted AI gateway written in Go that unifies access to multiple LLM providers and channels behind one endpoint. It manages API keys and subscription accounts across channels, with scheduling, failover, request logging, and usage tracking built in.
When you juggle many API keys or provider accounts across OpenAI, Anthropic, Gemini, and similar services, it gives you one place to route, balance, and monitor those requests.
Use it to
- Pool multiple API keys behind one endpoint with load balancing
- Set up failover between providers or channels
- Track request logs and per-key usage
- Self-host a gateway in front of subscription accounts
- Serve Claude Code or Codex clients through managed channels
For Developers self-hosting LLM infrastructure with multiple keys or providers
- Role
- other
- Language
- Go
- Licence
- MIT
- Forks
- 732
- Open issues
- 12
- Last push
- 2026-09-17
- Latest release
- v0.1.1 · 2025-06-13
topicsai-gatewayllmload-balancingfailoverself-hostedgo