Open-source LLM observability: log cost, latency, tokens, and agent traces for every OpenAI, Anthropic, Gemini, or Ollama call with a one-line baseURL swap. Self-hostable, MIT.

Spanlens is an open-source (MIT) LLM observability platform that lets developers monitor every call their application makes to OpenAI, Anthropic, Gemini, Mistral, OpenRouter, Azure OpenAI, or a local Ollama model. Integration takes one line: swap your client's baseURL to the Spanlens proxy, or run "npx @spanlens/cli init" and the wizard rewrites your code automatically. From that moment, every request is recorded with its model, token counts, latency, cost, and full prompt and response body, with streaming responses reconstructed automatically.

The dashboard turns that raw log into operational insight. Cost tracking breaks spend down per request, per model, and per end user, and parses prompt-cache tokens separately so you see real cache savings rather than sticker price. Agent tracing visualizes multi-step workflows as Gantt waterfalls and node-and-edge graphs, highlighting the critical path so you can find the slowest dependency chain in a fan-out. Anomaly detection flags 3-sigma deviations in latency, cost, or error rate against a rolling 7-day baseline with root-cause hints. Alerts on budget, error rate, and p95 latency are delivered to Email, Slack, or Discord.

Spanlens goes beyond passive logging. A regex-based PII and prompt-injection scanner inspects request and response bodies and can block injections at the proxy. The savings engine spots calls that match a cheaper model's profile (for example, a gpt-4o call that looks like a classification task) and estimates the monthly saving from switching. Prompt versioning with A/B experiments compares versions on latency, cost, and error rate using Welch's t-test for statistical significance, and an LLM-as-judge evaluation framework (judge with OpenAI, Anthropic, or Gemini) scores outputs against rubric anchors, with human agreement measured by Pearson r or Cohen's kappa. Reusable datasets power offline evals and regression checks.

The platform is built for teams: multiple workspaces and projects, role-based access, email invitations, and a full audit log. It also speaks open standards, accepting OpenTelemetry OTLP trace exports and shipping a documented REST API with an OpenAPI spec and Swagger UI. A bundled MCP server lets you query your observability data directly from Cursor, Claude Desktop, or Continue.

Because it is MIT-licensed, Spanlens can be self-hosted with a single Docker command, keeping all data inside your own infrastructure. The hosted service at spanlens.io offers a free tier of 50K requests per month with a hard cap, so a runaway loop can never surprise you with a bill, plus paid plans that use fair, authorized overage instead of dropping requests. Spanlens is the lightweight alternative to heavier or recently-acquired tools like Langfuse and Helicone: the essential 20 percent of features that cover 80 percent of real production needs, with a two-minute setup and pricing that does not punish growth.