# Weflayr Docs

> LLM observability platform: SDK and API reference

- [Weflayr Documentation](https://docs.weflayr.com/index.md): Weflayr helps you continuously optimize your margin through AI cost reduction. It joins real revenue to real cost, per customer and per feature, for instant insights, and automatically finds optimizations from your real usage. Integrate in minutes.
- [Quickstart](https://docs.weflayr.com/quickstart.md): Track your LLM calls, join them to your revenue, and start optimising, in 5 minutes.
- [Integrate with a coding agent](https://docs.weflayr.com/for-agents.md): Hand Weflayr's integration playbook to your coding agent and let it instrument your codebase.
- [Weflayr integration playbook](https://docs.weflayr.com/agents/integrate-weflayr.md): Step-by-step instructions for an AI coding agent integrating Weflayr into an application. Written to be read by the agent.
- [Technical documentation](https://docs.weflayr.com/technical.md): The SDK streams your LLM calls' telemetry to Weflayr. The dashboard's data can then be accessed and managed programmatically, through the API or MCP.
- [Track your LLM calls](https://docs.weflayr.com/track-your-llm-calls.md): Install the SDK, auto-instrument your AI provider and tag each call with a feature and customer.
- [Multiple AI providers (example)](https://docs.weflayr.com/examples/multi-provider.md): A single app that calls both OpenAI and Anthropic.
- [AI Framework (example)](https://docs.weflayr.com/examples/ai-framework.md): Instrument the provider your framework calls, and every call the chain makes is captured.
- [Upload revenue](https://docs.weflayr.com/api/revenue.md): Manage revenue programmatically instead of uploading CSVs.
- [Upload key metrics](https://docs.weflayr.com/api/key-metrics.md): Manage the key-metric values of an internal project programmatically instead of uploading CSVs.
- [Upload customer tags](https://docs.weflayr.com/api/customer-tags.md): Manage customer tags programmatically instead of uploading CSVs.
- [Get costs](https://docs.weflayr.com/api/costs.md): Get cost, bucketed by period and optionally grouped by one dimension (customer/feature/model/provider).
- [Get margin](https://docs.weflayr.com/api/margin.md): Get cost and revenue, bucketed by period and optionally grouped by customer, alongside the resulting margin and risk status.
- [Get performance](https://docs.weflayr.com/api/performance.md): Get cost, bucketed by period and optionally grouped by user, alongside each of the project's key metrics' value and cost-per-unit for the same period.
- [Get metadata](https://docs.weflayr.com/api/metadata.md): Get the project's identity, and the distinct dimensions (models, providers, features, tags, customers) it has seen.
- [Weflayr MCP](https://docs.weflayr.com/mcp-server.md): Access your Weflayr's data through MCP.
- [Python SDK Changelog](https://docs.weflayr.com/changelog/python.md)
- [Node.js SDK Changelog](https://docs.weflayr.com/changelog/node.md)
- [AI Gateway Changelog](https://docs.weflayr.com/changelog/ai-gateway.md)
- [Platform documentation](https://docs.weflayr.com/product.md): Weflayr allows you to analyse your AI costs & revenue and optimise it.
- [Cost & revenue tracking](https://docs.weflayr.com/product/cost-and-revenue-tracking.md): Once your app is instrumented with the SDK, every AI call streams to your dashboard.
- [Overview](https://docs.weflayr.com/product/overview.md): The Overview page is your daily pulse. It helps you spot if anything unusual is happening with you spend today or yesterday in one glance.
- [Cost: Cost per Feature](https://docs.weflayr.com/product/cost.md): Break your AI spend down per feature.
- [Margin: P&L per Customer](https://docs.weflayr.com/product/margin.md): The Margin page shows your revenue against your AI cost per customer.
- [AI Cost Optimisation](https://docs.weflayr.com/product/cost-optimisation.md): The Optimisation engine finds ways to cut your AI cost without changing what your app does.
- [Model Benchmarking](https://docs.weflayr.com/product/model-benchmark.md): The Model Benchmarking tries to find cheaper model that performs as well as the one you run today.
- [Prompt caching optimisation](https://docs.weflayr.com/product/prompt-caching.md): Providers charge less for tokens at the start of a prompt they have already seen.
- [Prompt caching analyzer](https://docs.weflayr.com/product/caching-analyzer.md): Providers do not always perfectly follow what your caching settings asked for, so your LLM calls may cost more than what you expect. This page tracks those misses.
- [Internal projects](https://docs.weflayr.com/internal-projects.md): Track an internal AI tool's cost against the key metrics it drives, instead of revenue.
- [Configure Cost & revenue](https://docs.weflayr.com/configure/introduction.md): The Costs & revenue page in the dashboard sidebar is where you feed Weflayr everything that is not an AI call.
- [Sync Revenue](https://docs.weflayr.com/configure/sync-revenue.md): The Margin page needs your revenue per customer.
- [Customer tags](https://docs.weflayr.com/configure/customer-tags.md): Tags add your own dimensions to customers (e.g. segment).
- [Free AI credits](https://docs.weflayr.com/configure/free-ai-credits.md): If a provider granted you free credits, your invoiced cost is lower than list prices until they run out.

## Optional

- [For agents](https://docs.weflayr.com/for-agents)
- [Talk to us](https://calendly.com/noe-weflayr/30min)