A production runtime for agents, batteries included.
FastAgentic gives you everything a serious agent platform needs — and nothing you'll immediately want to tear out.
Built for the way agents ship to production.
FastAgentic sits between your agent framework and your users — handling protocols, durability, governance, and telemetry so your team can focus on reasoning, not plumbing.
One decorator, three protocols
Define an agent endpoint once and expose it simultaneously as REST, MCP (Model Context Protocol), and A2A (Agent-to-Agent). Schemas stay in lock-step automatically.
Framework-agnostic adapters
Bring PydanticAI, LangGraph, CrewAI, LangChain — or a custom Runnable. Swap frameworks without rewriting deployment, auth, or observability.
Durable checkpoints
StepTracker and run_opaque cache workflow progress to Redis, Postgres, or S3. Resume after crashes without external orchestrators.
Policy & cost control
Budget caps, per-tenant rate limits, RBAC, and PII masking baked in. Stop runaway LLM bills before they start.
Streaming-first
SSE, WebSocket, and MCP events out of the box. Token streaming, tool calls, and intermediate steps — all with zero boilerplate.
Production observability
OpenTelemetry, Langfuse, Portkey, and Datadog integrations. Structured logs, per-run cost tracking, and audit trails ready on day one.
Protocols
- REST
- Standard FastAPI routes with streaming SSE and WebSocket upgrades.
- MCP
- Model Context Protocol 2025-11-25. Auto-generated tool, resource, and prompt surfaces.
- A2A
- Agent-to-Agent v0.3 skills, negotiations, and multi-agent message passing.
- OpenAPI 3.1
- Fully-typed OpenAPI schemas shared across all three surfaces.
Framework adapters
- PydanticAI
- First-class adapter, zero-config integration.
- LangGraph
- State machines wrapped with StepTracker for resumable runs.
- CrewAI
- Multi-agent crews exposed as single endpoints.
- LangChain
- Runnables, chains, and agents via a single adapter.
- Custom
- Implement the Runnable interface for anything not covered.
Durability & resumption
- StepTracker
- Fine-grained checkpoints per workflow step.
- run_opaque
- Wrap any agent unchanged for resumable execution.
- Storage backends
- Redis, Postgres, S3, or custom adapter.
- Crash-safe
- Resume mid-run after process restarts with no external orchestrator.
Governance & safety
- OAuth2 / OIDC
- Built-in auth with your existing IdP.
- Cost tracker
- Per-run, per-user, per-tenant token + pricing.
- Budget enforcement
- Hard caps that stop runaway LLM spend.
- Policy engine
- RBAC, scope checks, and custom policies.
- PII masking
- Automatic detection and redaction.
- Audit logs
- Immutable trail of every agent decision.
Observability
- OpenTelemetry
- Distributed traces across every step.
- Langfuse
- Native LLM observability.
- Portkey
- Multi-provider gateway support.
- Datadog
- APM and log integration.
- Structured logs
- JSON with full request context.
Developer experience
- CLI (fastagentic)
- new · run · agent chat · inspect · migrate.
- Hot reload
- Fast iteration during development.
- Typed schemas
- Pydantic 2 end-to-end.
- Examples
- 9 reference apps covering every adapter.
Need FastAPI, LangGraph, or agent platform expertise?
Neul Labs — the team behind FastAgentic — takes on a limited number of consulting engagements each quarter. We help teams ship agents to production, fix broken LangGraph pipelines, and design governance for multi-tenant LLM platforms.