# Letta

**Canonical:** https://apis.io/providers/letta/  
**Website:** https://www.letta.com/  
**APIs profiled:** 41

Letta (formerly MemGPT) is a stateful AI agents platform built around long-term memory, tool execution, and multi-agent coordination. The Letta REST API exposes 239 endpoints across 36 public resource categories — agents, memory blocks, archival memory, sources (RAG), custom tools (sandboxed/client-side/MCP), MCP servers, multi-agent groups, identities, runs, scheduled messages, and a streaming voice-mode endpoint. Open-source server (Apache-2.0, 22.9k+ stars) is available on GitHub; Letta Cloud is the managed offering; the Agent Development Environment (ADE) provides a web UI for inspecting context windows, memory blocks, and run history. Python and TypeScript SDKs ship alongside the REST API, and the open `.af` agent file format lets agents migrate between deployments.

## Kin Score — 45.3 / 100 (developing)

Scored 2026-08-20 under rubric 0.12.0. Trend: flat (+0.0 from 45.3).

| Facet | Score |
|---|---|
| Discoverability | 75.9 |
| Contract Quality | 83.5 |
| Governance | 26.5 |
| Contract Governance | 26.5 |
| Operational Transparency | 7.9 |
| Developer Ergonomics | 23.8 |
| Commercial Clarity | 39.5 |
| Access Clarity | 39.5 |

## Agent readiness — 45.6 (agent-ready)

| Dimension | Value |
|---|---|
| Spec Presence | yes |
| Agentic Access | derived |
| Reversibility Documented | documented |
| MCP Server | no |
| Auth Clarity | yes |
| Idempotency | no |
| Error Semantics | verified |
| OpenAPI Examples | partial |
| Rate Limit Signal | documented |
| Event Surface Described | derived |
| Agent Skills | no |
| Well Known Catalog | no |
| Consent Identity | no |
| Agent Card | no |
| Dry Run Mode | yes |

## Access

Freemium · Self-serve signup — onboarding: self-serve, pricing: freemium, trial: no (confidence: high).

## APIs (41)

- **Letta Code** — Letta's memory-first coding agent — a CLI plus a desktop app (and an action for GitHub repos). "The memory-first coding agent that remembers and learns." Builds on the Letta Cod...
- **Letta Admin API** — Administrative operations (org / billing / policy).
- **Letta Agents API** — Create, list, modify, export, and delete stateful agents with memory and tools.
- **Letta Archives API** — Manage archival memory archives and passages (RAG-style long-term store).
- **Letta Blocks API** — Manage in-context memory blocks (core memory) shared across agents.
- **Letta Chat API** — Chat operations.
- **Letta Client-Side Access Tokens API** — Mint short-lived client-side access tokens for browser/mobile use.
- **Letta Conversations API** — Conversational session threads anchored to an agent.
- **Letta Device Storage API** — Per-device storage for the Letta Code SDK and ADE.
- **Letta Embeddings API** — List configured embedding models.
- **Letta Environments API** — Environment variables and configuration scoped to agents/tools.
- **Letta Feeds API** — Subscribed feeds that drive agents.
- **Letta Folders API** — Organize agents and resources into folders.
- **Letta Groups API** — Coordinate multi-agent groups (supervisor/worker, round-robin, sleep-time).
- **Letta Health API** — Service health/readiness endpoint.
- **Letta Identities API** — Map external user identities to agents and memory contexts.
- **Letta Internal Agents API** — Internal Agents operations.
- **Letta Internal Blocks API** — Internal Blocks operations.
- **Letta Internal Runs API** — Internal Runs operations.
- **Letta Internal Templates API** — Internal Templates operations.
- **Letta Jobs API** — Manage long-running background jobs (e.g., file ingestion).
- **Letta LLMs API** — List available LLM models and configurations.
- **Letta MCP Servers API** — Register external Model Context Protocol servers and expose their tools to agents.
- **Letta Memory Files API** — Memory-mapped file resources attached to agents.
- **Letta Messages API** — Send messages, list message history, search, and batch operations.
- **Letta Metadata API** — Service/runtime metadata.
- **Letta Models API** — List available chat and embedding models exposed by configured providers.
- **Letta Passages API** — Individual archival memory passages (chunked context).
- **Letta Pipelines API** — Composable agent pipelines.
- **Letta Projects API** — Project-level scoping for agents and resources.
- **Letta Providers API** — Configure LLM provider credentials (OpenAI, Anthropic, Azure, Ollama, etc.).
- **Letta Runs API** — Inspect and manage agent runs (executions of an agent in response to a message).
- **Letta Sandboxes API** — Manage code-execution sandboxes used by sandbox tools.
- **Letta Scheduled Messages API** — Schedule one-time or recurring messages to drive agents over time.
- **Letta Sources API** — Manage data sources (files, URLs) used to populate agent archival memory.
- **Letta Steps API** — Per-step traces of an agent run.
- **Letta Tag API** — Tag-based filtering metadata.
- **Letta Telemetry API** — Run-step traces and observability data.
- **Letta Templates API** — Reusable agent templates for instantiation.
- **Letta Tools API** — Register and execute custom tools (sandboxed Python, client-side, MCP).
- …and 1 more, listed in full on the page.

## Agentic access (1)

- **Letta Agentic Access** — 604 operations · 346 acting · 8 human-in-the-loop

## Security (2)

- **Letta Authentication** — http · 1 scheme
- **Letta Domain Security** — TLSv1.3 · HSTS · DMARC

## Plans (1)

- **Letta Plans Pricing**

## Tags

Artificial Intelligence, Agents, Stateful Agents, Memory, MemGPT, Continual Learning, MCP, Multi-Agent, RAG, Open-Source

---

Profiled by [API Evangelist](https://apievangelist.com) and published on [APIs.io](https://apis.io/providers/letta/). Scores are computed from the provider's own public artifacts under a published rubric.