Documentation Index
Fetch the complete documentation index at: https://docs.promptlayer.com/llms.txt
Use this file to discover all available pages before exploring further.
Prompt Registry
Give teams a dedicated place to experiment with prompts and manage versions, separate from your application code.
Evaluations
Test prompts, define quality dimensions and scoring logic, and track scores across time.
Dataset Management
Build, organize, and reuse datasets for evaluation, regression testing, and iteration.
Observability
Review inference usage, cost, and latency. Inspect individual request/response pairs alongside full AI-application traces.
Workflows
Build complex, multi-step processes by chaining prompts and logical operations together.
Features
Release Labels
Promote tested prompt versions across environments without changing application code.
AB Testing
Compare prompt variants in production and route traffic based on measured performance.
Analytics
Track usage, quality, and outcome trends as prompts and workflows evolve over time.
Role-Based Access Control
Manage workspace permissions with granular roles for engineers, operators, and reviewers.
Single Sign-On
Enable secure team access with enterprise identity providers and centralized login flows.
Self-Hosting
Deploy PromptLayer in your own environment when security or compliance requires it.
Resources
API Reference
Explore the REST API surface for prompt execution, datasets, and workflows.
MCP
Documentation for model context protocol integrations, tools, and connected workflows.
Migration Guides
Move hard-coded prompts and ad hoc eval scripts into a managed PromptLayer workflow.
Webhooks
Explore supported events and payload details.
SDKs
Official client libraries, examples, and package references for supported runtimes.
OpenTelemetry Integrations
Connect PromptLayer with model providers, tracing pipelines, and internal tooling.
Changelog
Product updates covering PromptLayer platform improvements and launches.
Blog
Engineering and product insights from the PromptLayer team.