Manifest

What is Manifest?

Manifest is an open source LLM router and gateway for AI agents and AI apps. You connect the providers you already use, including pay-per-token API keys, subscriptions such as ChatGPT or Claude, local models and custom endpoints, and send every request through one URL that speaks both the OpenAI and Anthropic APIs. Model choice per request is yours, and you can add fallbacks so a rate limit or retired model still returns a working answer, cap spending per key, and read the full request and response body of every call. Autofix repairs malformed requests and resends them instead of letting them fail. It runs as Manifest Cloud or self-hosted with Docker under the MIT licence, and Manifest does not mark up tokens, since you pay providers directly. Engineers and platform teams use it to keep Claude Code, the OpenAI SDK or the Vercel AI SDK working behind a single base URL. Self-hosting is free; the Free cloud plan routes 10,000 requests a month until November 1, 2026, and Manifest Cloud is $19 a month from November 2.

Use cases

• Route agent requests through one endpoint
• Fall back to another model on failure
• Cap LLM spend per key or engineer
• Run local models alongside cloud ones
• Repair malformed LLM requests automatically
• Inspect full request and response logs

Who it's for

• AI engineers
• Platform teams
• Startups
• Enterprise IT teams

Why use Manifest

• One base URL change lets existing tools like Claude Code or the OpenAI SDK use many providers.
• Fallbacks and Autofix keep agents working when a provider errors or a request is malformed.
• Per-key budgets and cost breakdowns show who spent what before the invoice arrives.
• You bring your own provider keys and Manifest adds no token markup.
• The MIT-licensed gateway can run on your own servers with no request limit.

Manifest features

• One endpoint compatible with OpenAI and Anthropic APIs
• API key, subscription, local and custom providers
• Model fallbacks across providers
• Autofix for broken requests
• Budget limits, rate limits and alerts
• Full request and response body logs
• Cost breakdown by agent, key and provider
• Ollama, LM Studio and llama.cpp support
• SSO / SAML and audit logs on Enterprise
• Self-hosting with Docker under the MIT licence

Manifest integrations

• OpenAI
• Anthropic
• Ollama
• LM Studio
• llama.cpp
• Claude Code
• Vercel AI SDK
• GitHub Copilot
• Railway

How to get started with Manifest

1. Sign up for Manifest Cloud or run the gateway yourself with Docker.
2. Connect your providers: API keys, subscriptions, local models or custom endpoints.
3. Point your agent or SDK at the Manifest URL, which speaks the OpenAI and Anthropic APIs.
4. Pick models per route, add fallbacks and set budget limits and alerts.
5. Watch spend, failures and full request logs in the dashboard.

Manifest limitations

• Manifest Cloud cannot reach local models running on your machine
• On the Free cloud plan, routing pauses after 10,000 requests in a month
• From November 2, 2026 a Free cloud workspace stops routing requests
• Pro is a single-user plan; seats and SSO need Enterprise

Manifest pricing

Manifest is open source, so you can use and self-host it for free. Paid plans start at $19 per month, according to the official pricing page on 24 Oct 2026.

Manifest plans

• Free

  $0 /month

  • Cloud or self-hosted with Docker, MIT licence
  • Unlimited agents
  • 10,000 routed requests per month on Cloud
  • All providers
  • 7-day dashboard retention
  • Autofix and budget alerts

• Pro

  $19 /month

  • Hosted on Manifest Cloud
  • Unlimited routed requests
  • 365 days dashboard retention
  • Basic support

• Enterprise

  Custom

  • On-premise or VPC deployment
  • Multiple seats and team management
  • SSO / SAML and audit logs
  • Contractual uptime SLA
  • HIPAA and custom BAAs
  • Dedicated support

Plans as listed on the official pricing page on 24 Oct 2026. Prices can change and may vary by country or exclude tax. Check the official pricing page.

Manifest alternatives

The main Manifest alternatives listed on Debuly are LiteLLM, Portkey, OpenRouter, Helicone, Bifrost and Blackbox AI. The table compares pricing and platforms; order is A–Z.

1. 

   Bifrost

   Open-source AI gateway from Maxim with one OpenAI-compatible API across providers, failover, budgets and an MCP gateway.

   Developer platformsAutomation and agents
   Open sourceVisit Bifrost (official site)
2. 

   Blackbox AI

   Inference platform with one OpenAI-compatible endpoint for 300+ open and closed models, encrypted connections, and dedicated enterprise deployments.

   Developer platformsCodingAutomation and agents
   Contact salesVisit Blackbox AI (official site)
3. 

   Helicone

   Open-source LLM observability and AI gateway that logs requests, tracks cost and latency, and routes to 100+ models through one API.

   Developer platformsCoding
   FreemiumVisit Helicone (official site)
4. 

   LiteLLM

   Open-source AI gateway and Python SDK that gives one OpenAI-format API to 100+ LLM providers, with keys, budgets and cost tracking.

   Developer platformsCodingAutomation and agents
   Open sourceVisit LiteLLM (official site)
5. 

   OpenRouter

   Unified OpenAI-compatible API for 500+ AI models with provider fallbacks, pay-as-you-go credits and a free tier

   Developer platformsChat assistantsAutomation and agents
   FreemiumVisit OpenRouter (official site)
6. 

   Portkey

   AI gateway with observability, guardrails, caching and prompt management for teams running LLM apps in production.

   Developer platformsAutomation and agents
   FreemiumVisit Portkey (official site)

Manifest FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from manifest.build · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.