fal
Media
Fast image, video, and audio generation on serverless GPUs with a queue API and webhooks
OpenAPIllms.txt
API keypay-per-second
A directory of services AI agents can actually use — APIs, tools and merchants with machine-readable interfaces and agent payment protocols. Badges are probe-backed, never self-asserted.
Models
Gemini multimodal models via Google AI Studio and Vertex AI, with function calling and MCP support
Models
Serverless and dedicated inference for open models, plus a hosted MCP endpoint over Hub tools
Observability
Open-source LLM and agent observability — traces, evals, prompt management, and cost tracking
Platforms
Hosted runtime for stateful agent graphs with persistence, streaming, and human-in-the-loop
Observability
Tracing, evals, and prompt management with the deepest LangChain and LangGraph integration
Memory
Stateful agent runtime where memory is the agent — self-editing core memory, archival recall, and a REST API
Platforms
Issue tracking and project data over GraphQL and a remote MCP server agents can drive
Models
Open-source proxy and SDK exposing 100+ model providers behind one OpenAI-compatible API
Messaging
Open-source realtime media infrastructure with the Agents framework, SIP, and phone numbers
Payments
Open HTTP payment standard from Stripe and Tempo covering one-shot, recurring, and streaming agent payments across cards, stablecoins, and Lightning
Memory
Managed memory layer that extracts, dedupes, and retrieves long-term memories per user, agent, or session
Compute
Serverless GPU and CPU compute defined in Python, with sandboxes and pay-per-second MPP billing
Platforms
Open-source workflow automation with a public API, webhook triggers, and MCP nodes
Memory
Serverless Postgres with instant branching, autoscaling, and a remote MCP server
Platforms
Open catalog of MCP servers with a public API agents and clients can query for installable tools
Models
GPT-5 series models, embeddings, and the Responses API with native MCP and tool use
Models
One OpenAI-compatible endpoint in front of hundreds of hosted models, with automatic fallback and per-key spend limits
Search
Search and deep-research APIs built for agents, with tiered latency and citation-grade results
Same query as JSON: /api/v1/services?page=2
