Skip to main content

Take control of your OpenClaw costs

Smart LLM routing and observability platform that cuts AI costs up to 70% with real-time tracking, intelligent model selection, and comprehensive telemetry.

Get StartedAPI Reference

Quick start

Get up and running with Manifest in minutes

1

Install the plugin

Install Manifest as an OpenClaw plugin with a single command:
openclaw plugins install manifest
The plugin hooks into your OpenClaw gateway and starts routing and tracking all LLM requests automatically.
2

Choose your mode

Manifest runs in two modes — cloud (default) or local:
Sign up at app.manifest.build to get your API key, then configure:
openclaw config set plugins.entries.manifest.config.apiKey "mnfst_YOUR_KEY"
openclaw gateway restart
Your telemetry flows to the cloud dashboard with secure blind proxy architecture.
3

View your dashboard

Your agent telemetry starts flowing immediately. Open the dashboard to see:
  • Real-time cost and token tracking
  • Message logs with full conversation history
  • Model usage breakdowns across 300+ supported models
  • Smart routing decisions and scoring insights
In local mode, the dashboard shows an orange Dev badge. In cloud mode, access your dashboard at app.manifest.build.

Explore by topic

Dive deeper into Manifest’s features and capabilities

Core concepts

Understand how Manifest routes requests, tracks costs, and integrates with OpenClaw

Dashboard

Explore real-time analytics, cost tracking, and message logs

LLM routing

Learn how the 23-dimension scoring algorithm selects optimal models

Observability

OpenTelemetry native with OTLP traces, metrics, and logs

Multi-agent

Manage multiple agents with separate tracking and cost allocation

Self-hosting

Deploy your own Manifest instance with PostgreSQL and custom config

Features

Everything you need to optimize your AI costs

Smart routing

23-dimension scoring algorithm that analyzes every query and routes to the most cost-effective model in under 2ms.

Learn more

Cost tracking

Real-time dashboards show tokens and costs per model, agent, and time period with pricing database covering 300+ models.

Learn more

OTLP native

Standard OpenTelemetry ingestion with traces, metrics, and logs for full visibility into agent behavior.

Learn more

Alerts & notifications

Set usage thresholds and receive email alerts when your consumption exceeds limits you define.

Learn more

Resources

Learn more and connect with the community

GitHub repository

Star the repo, report issues, and contribute to the open-source project

Discord community

Join the conversation, get help, and share feedback with the team

Contributing guide

Learn how to contribute code, report bugs, and improve the documentation

Architecture deep dive

Explore the technical architecture and design decisions

Ready to optimize your AI costs?

Install Manifest in minutes and start tracking costs, routing requests intelligently, and gaining full observability into your agents.

Get started now