LibreChat
Open-source AI chat platform supporting multiple model providers and tools.
LibreChat is becoming a self-hosted agent platform: skills, subagents, and frontier models.
◆Recent moves
- 26d ago
v0.8.7: skill authoring, agent marketplace, native Anthropic + GPT-5.5
⚡ SPARKThe largest entry in the window pushes the agent platform forward - skill file-authoring tools, GitHub skill sync, an agent marketplace in the model selector, and model-spec subagents - while adding native Anthropic endpoints, Claude Fable 5, and GPT-5.5. It extends exactly the arc 0.8.6 started.
View source ↗ - 27d ago
chart-2.0.6
A Helm chart version bump to 2.0.6 with no user-facing change - deployment packaging keeping pace with the app releases.
View source ↗ - 1mo ago
chart-2.0.4: 🪪 fix: Add Admin Panel SSO URL Config (#13220)
A chart fix clarifying admin-panel SSO URL configuration and removing a duplicate env var. Operational plumbing for self-hosters, not a product capability change.
View source ↗ - 2mo ago
chart-2.0.3
A chart release that mostly aggregates the v0.8.6-rc1 app changelog and a long new-contributor list - packaging, not product.
View source ↗ - 2mo ago
v0.8.6: Agent Skills and subagents arrive
⚡ SPARKThe release that set the current direction - Agent Skills (SKILL.md bundles of instructions, scripts, and tool permissions) and subagents that let an agent call other agents as tools. Code-execution artifacts and CloudFront delivery round it out.
View source ↗ - 3mo ago
v0.8.5: multi-tenant admin APIs and RBAC
Enterprise groundwork - admin APIs for users, groups, roles, and grants, DB-backed per-principal config overrides, tenant-scoped app config, a 3-tier MCP architecture, and a redesigned tool-call UI. The plumbing that later releases' agent features depend on.
View source ↗