[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-82914":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":27,"readmeContent":28,"aiSummary":29,"trendingCount":16,"starSnapshotCount":16,"syncStatus":15,"lastSyncTime":30,"discoverSource":31},82914,"litellm-rust","LiteLLM-Labs\u002Flitellm-rust","LiteLLM-Labs","A minimal Rust gateway built for coding agents (LiteLLM-Compatible)","",null,"TypeScript",129,15,56,2,0,1,22,72,6,63.81,"MIT License",false,"main",true,[],"2026-06-12 04:01:39","# litellm-rust\na simple and blazing fast LiteLLM-compatible ai gateway for coding agents (Claude Code, Codex, Hermes, etc.) \n\n[![Discord](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-Chat-5865F2?logo=discord&logoColor=white)](https:\u002F\u002Fdiscord.gg\u002FNkxw3rm3EE)\n\n![LLM Gateway Proxy Overhead](docs\u002Fbenchmark.png)\n\n## Usage with Claude code\n\n```bash\nlite claude\n```\n\nThe first run prompts for your LiteLLM URL and API key, saves them to\n`~\u002F.config\u002Flite\u002Fclaude.env`, and starts Claude Code with:\n\n```bash\nANTHROPIC_BASE_URL=\"https:\u002F\u002Fyour-litellm-rust-server.com\"\nANTHROPIC_AUTH_TOKEN=\"$LITELLM_API_KEY\"\n```\n\nArguments after `lite claude` are forwarded to Claude Code:\n\n```bash\nlite claude --help\nlite claude --model claude-sonnet-4-5\n```\n\nRun `lite claude --reset` to ignore saved settings and enter them again.\n\n## Quickstart\n\nlitellm-rust is compatible with your existing litellm config.yaml and DB. \n\n```yaml\nmodel_list:\n\t- model_name: anthropic\u002F*\n\t\tlitellm_params:\n\t\t\tmodel: anthropic\u002F*\n\t\t\tapi_key: os.environ\u002FANTHROPIC_API_KEY\n\n\ngeneral_settings:\n\tmaster_key: os.environ\u002FMASTER_KEY\n\tsandbox_choice: \"e2b\" # can be either \"e2b\" or \"daytona\"  \n\te2b_sandbox_params:\n\t\te2b_api_key: os.environ\u002FE2B_API_KEY\n\t\te2b_template: \"litellm-4gb\"\n```\n\n```\n$ litellm-rust --config \u002Fapp\u002Fconfig.yaml\n```\n\n## Routes\n\n```txt\nPOST \u002Fmessages\nPOST \u002Fresponses\nPOST \u002Frealtime\nPOST \u002Faudio\n```\n\n## Providers\n- OpenAI\n- Azure OpenAI\n- VertexAI\n- Bedrock\n\n## Codebase map\n\nEntry points and what runs at startup:\n\n- **`src\u002Fmain.rs`** — binary entry point. Parses CLI args, loads `config.yaml`, builds the HTTP client, calls `model_prices::load()`, then wires everything into `AppState` and starts the server.\n- **`src\u002Fmodel_prices.rs`** — fetches the LiteLLM model cost\u002Fcapability map from upstream at startup; falls back to the embedded `model_prices_backup.json` snapshot if the network is unavailable. Returns a `ModelCostMap`; `main.rs` stores it on `AppState`. Override the URL with `LITELLM_MODEL_COST_MAP_URL`.\n- **`src\u002Ferrors.rs`** — typed error enum. All error variants map to HTTP status + JSON body in one place.\n\nSubsystems:\n\n- **`src\u002Fhttp\u002F`** — HTTP layer only. Route registration, auth, body extraction, response shaping. No business logic.\n- **`src\u002Fproviders\u002F`** — provider registry, per-provider request\u002Fresponse transformation, model router (maps model name → deployment + handler).\n- **`src\u002Fproxy\u002F`** — config loading, master-key auth, `AppState`.\n- **`src\u002Fcli\u002F`** — `lite claude` wizard: credential storage, model selector, Claude Code launcher.\n\n## Coding standards\n\nSee [CODING_STANDARDS.md](CODING_STANDARDS.md).\n","litellm-rust 是一个为编码代理（如Claude Code、Codex等）设计的轻量级Rust网关。它提供了与LiteLLM兼容的API接口，支持快速且高效的AI服务接入。该项目的核心功能包括通过简单的命令行工具来配置和启动编码代理，同时支持多种主流AI服务提供商，如OpenAI、Azure OpenAI等。litellm-rust特别适用于需要高效处理AI请求、追求低延迟响应以及希望简化AI集成流程的应用场景。其简洁的设计和强大的兼容性使其成为开发人员构建基于AI的应用程序的理想选择。","2026-06-11 04:09:38","CREATED_QUERY"]