Add new SambaNova models (DeepSeek-V3.1-cb, DeepSeek-V3.1-Terminus, DeepSeek-V3.2, Qwen3-235B, Llama-3.3-Swallow-70B-Instruct-v0.4, ALLaM-7B-Instruct-preview)
and update existing models to reflect current pricing and capabilities from SambaNova Cloud.
Rename DeepSeek-R1 to DeepSeek-R1-0528 to match versioned model name,
expand context windows to 128k where supported,
and retain legacy models for backward compatibility with custom environments.
Implements X-Integration-Source header for SambaNova chat and embedding endpoints to enable analytics tracking.
Emulating same behavior as Sambanova SDK, supports override via integration_source parameter or SAMBANOVA_INTEGRATION_SOURCE env var.
with default value "litellm".
* feat(ui): Add team-alias column to Models Health Status UI
- Added Team Alias column to the Models Health Status table
- Updated HealthCheckComponent to accept teams prop
- Updated health_check_columns to display team alias based on team_id
- Falls back to team_id if team alias not found, or shows '-' if no team
- Updated parent components to pass teams data to HealthCheckComponent
Co-authored-by: ishaan <ishaan@berri.ai>
* Update ui/litellm-dashboard/src/components/model_dashboard/health_check_columns.tsx
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: ishaan <ishaan@berri.ai>
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
* docs: add Prisma migration troubleshooting guide
Add troubleshooting documentation for common Prisma migration errors
encountered when upgrading/downgrading LiteLLM proxy versions.
Covers:
- 'relation does not exist' errors after version rollback
- Blocked migrations from previous failures
- Migration state mismatch after version rollback
- General tips for prisma migrate resolve, db push, and migrate deploy
* docs: simplify prisma migration troubleshooting - focus on delete + restart
Add the following SCIM v2 discovery endpoints per RFC 7643/7644:
- GET /scim/v2 - Base resource discovery (ListResponse of ResourceTypes)
- GET /scim/v2/ResourceTypes - List all supported resource types
- GET /scim/v2/ResourceTypes/{id} - Get a specific resource type (User/Group)
- GET /scim/v2/Schemas - List all supported schemas
- GET /scim/v2/Schemas/{uri} - Get a specific schema by URI
These endpoints are required by identity providers (Okta, Azure AD, etc.)
for SCIM resource discovery. Previously, GET /scim/v2 returned 404.
Also adds SCIMResourceType, SCIMSchema, and SCIMSchemaAttribute Pydantic
models to the SCIM types module.
Fixes#20295
* fix: strip stale mcp-session-id header to prevent 'Session not found' error loop
When VSCode reconnects to LiteLLM's MCP endpoint after a reload, it sends
a stale mcp-session-id header. The session was already cleaned up, causing
a 404 'Session not found' error. VSCode retries with the same stale ID,
creating an infinite error loop.
Before forwarding requests to the StreamableHTTP session manager, check if
the mcp-session-id header references a valid session. If the session doesn't
exist, strip the header so a new session is created automatically.
Fixes#20292
* refactor: extract stale session handling into _strip_stale_mcp_session_header helper
Add moonshot/kimi-k2.5 model with:
- Input cost: $0.60/M tokens (6e-07)
- Output cost: $3.00/M tokens (3e-06)
- Cache read cost: $0.10/M tokens (1e-07)
- 256K context window
- Vision, function calling, tool choice, web search support
Reference: https://huggingface.co/moonshotai/Kimi-K2.5
Note: K2.5 thinking mode is controlled via API parameters, not a separate model ID.
Co-authored-by: krauckbot <krauckbot123@gmail.com>
