Manifest.build
LLM routing layer that connects AI agents and apps to over 15 model providers, adding fallbacks, cost controls, and auto-fix for bad requests.
About
Manifest.build sits between your AI application and the underlying language model providers, routing each request to the most appropriate model based on rules you define. It supports OpenAI, Anthropic, Google, Mistral, and more than a dozen other providers, as well as local models run through Ollama or LM Studio.
Key capabilities include automatic fallback to a backup provider if the primary fails, real-time cost dashboards, budget limits, and an auto-fix layer that corrects malformed or deprecated API calls before they reach the model. It is aimed at solo developers, teams running agent workflows, and users of coding tools who want to reduce infrastructure complexity. A free bring-your-own-key tier is available; advanced cloud features require a paid plan.
Key capabilities include automatic fallback to a backup provider if the primary fails, real-time cost dashboards, budget limits, and an auto-fix layer that corrects malformed or deprecated API calls before they reach the model. It is aimed at solo developers, teams running agent workflows, and users of coding tools who want to reduce infrastructure complexity. A free bring-your-own-key tier is available; advanced cloud features require a paid plan.