From 8eec2c69b74952be93190dd40da88156746c3600 Mon Sep 17 00:00:00 2001 From: Yuneng Jiang Date: Tue, 14 Apr 2026 15:58:13 -0700 Subject: [PATCH 1/4] [Docs] Add release notes for v1.83.3-stable and v1.83.7.rc.1 - Retitle existing v1.83.3 preview file to v1.83.3-stable (same commit) - Add new v1.83.7.rc.1 preview release notes - Update RELEASE_NOTES_GENERATION_INSTRUCTIONS runbook with guidance on resolving staging PRs to their underlying commits --- .../RELEASE_NOTES_GENERATION_INSTRUCTIONS.md | 9 + .../my-website/release_notes/v1.83.3/index.md | 10 +- .../release_notes/v1.83.7.rc.1/index.md | 210 ++++++++++++++++++ 3 files changed, 224 insertions(+), 5 deletions(-) create mode 100644 docs/my-website/release_notes/v1.83.7.rc.1/index.md diff --git a/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md b/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md index ab2cf334459..ef7b146fe07 100644 --- a/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md +++ b/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md @@ -9,6 +9,15 @@ This document provides comprehensive instructions for AI agents to generate rele 3. **Previous Version Commit Hash** - To compare model pricing changes 4. **Reference Release Notes** - Use recent stable releases (v1.76.3-stable, v1.77.2-stable) as templates for consistent formatting +### Resolving Staging PRs + +The GitHub release page (e.g. `https://github.com/BerriAI/litellm/releases/tag/v1.83.3-stable`) does **not** list the real changelog directly. The "What's Changed" section contains **staging PRs** that each bundle many individual commits/PRs. For example: + +- `Litellm oss staging 03 14 2026 by @RheagalFire in #23686` +- `Litellm ryan march 16 by @ryan-crabbe in #23822` + +To get the real changelog, you MUST click into each staging PR (e.g. `#23686`, `#23822`), open its **Commits** tab, and extract every underlying commit/PR (look for the `(#NNNNN)` suffix on commit titles). Those underlying PRs — not the staging PRs — are what get categorized in the release notes. Never treat a staging PR title as a single changelog entry. + ## Step-by-Step Process ### 1. Initial Setup and Analysis diff --git a/docs/my-website/release_notes/v1.83.3/index.md b/docs/my-website/release_notes/v1.83.3/index.md index bfa66b8fcc2..1eced9239c9 100644 --- a/docs/my-website/release_notes/v1.83.3/index.md +++ b/docs/my-website/release_notes/v1.83.3/index.md @@ -1,6 +1,6 @@ --- -title: "[Preview] v1.83.3.rc.1 - Introducing MCP Skills Marketplace" -slug: "v1-83-3-rc-1" +title: "v1.83.3-stable - Introducing MCP Skills Marketplace" +slug: "v1-83-3-stable" date: 2026-04-04T00:00:00 authors: - name: Krrish Dholakia @@ -38,14 +38,14 @@ import TabItem from '@theme/TabItem'; docker run \ -e STORE_MODEL_IN_DB=True \ -p 4000:4000 \ -docker.litellm.ai/berriai/litellm:main-v1.83.3.rc.1 +docker.litellm.ai/berriai/litellm:main-v1.83.3-stable ``` ```bash -pip install litellm==1.83.3rc1 +pip install litellm==1.83.3.post1 ``` @@ -233,4 +233,4 @@ MCP Toolsets let AI platform admins create curated subsets of tools from one or * @vanhtuan0409 made their first contribution in https://github.com/BerriAI/litellm/pull/24078 * @clfhhc made their first contribution in https://github.com/BerriAI/litellm/pull/24932 -**Full Changelog**: https://github.com/BerriAI/litellm/compare/v1.83.0-nightly...v1.83.3.rc.1 +**Full Changelog**: https://github.com/BerriAI/litellm/compare/v1.83.0-nightly...v1.83.3-stable diff --git a/docs/my-website/release_notes/v1.83.7.rc.1/index.md b/docs/my-website/release_notes/v1.83.7.rc.1/index.md new file mode 100644 index 00000000000..811b129d22b --- /dev/null +++ b/docs/my-website/release_notes/v1.83.7.rc.1/index.md @@ -0,0 +1,210 @@ +--- +title: "[Preview] v1.83.7.rc.1 - Per-User MCP OAuth, Team Spend Logs RBAC" +slug: "v1-83-7-rc-1" +date: 2026-04-12T00:00:00 +authors: + - name: Krrish Dholakia + title: CEO, LiteLLM + url: https://www.linkedin.com/in/krish-d/ + image_url: https://pbs.twimg.com/profile_images/1298587542745358340/DZv3Oj-h_400x400.jpg + - name: Ishaan Jaff + title: CTO, LiteLLM + url: https://www.linkedin.com/in/reffajnaahsi/ + image_url: https://pbs.twimg.com/profile_images/1613813310264340481/lz54oEiB_400x400.jpg + - name: Ryan Crabbe + title: Full Stack Engineer, LiteLLM + url: https://www.linkedin.com/in/ryan-crabbe-0b9687214 + image_url: https://media.licdn.com/dms/image/v2/D5603AQHt1t9Z4BJ6Gw/profile-displayphoto-shrink_400_400/profile-displayphoto-shrink_400_400/0/1724453682340?e=1772064000&v=beta&t=VXdmr13rsNB05wyA2F1TENOB5UuDHUZ0FCHTolNyR5M + - name: Yuneng Jiang + title: Senior Full Stack Engineer, LiteLLM + url: https://www.linkedin.com/in/yuneng-david-jiang-455676139/ + image_url: https://avatars.githubusercontent.com/u/171294688?v=4 + - name: Shivam Rawat + title: Forward Deployed Engineer, LiteLLM + url: https://linkedin.com/in/shivam-rawat-482937318 + image_url: https://github.com/shivamrawat1.png +hide_table_of_contents: false +--- + +## Deploy this version + +import Tabs from '@theme/Tabs'; +import TabItem from '@theme/TabItem'; + + + + +```bash +docker run \ +-e STORE_MODEL_IN_DB=True \ +-p 4000:4000 \ +docker.litellm.ai/berriai/litellm:main-v1.83.7.rc.1 +``` + + + + +```bash +pip install litellm==1.83.7rc1 +``` + + + + +:::warning + +**Breaking change — Prometheus latency histogram buckets reduced.** The default `LATENCY_BUCKETS` set has been reduced from 35 to 18 boundaries to lower Prometheus cardinality. Dashboards and PromQL queries that reference specific `le=` bucket values may stop matching. Review your alerts/dashboards before upgrading and use `LATENCY_BUCKETS` env override to restore the previous boundaries if needed — [PR #25527](https://github.com/BerriAI/litellm/pull/25527). + +::: + +## Key Highlights + +- **Per-User MCP OAuth Tokens** — [Each end-user can now hold their own OAuth tokens for interactive MCP server flows, isolating credentials across users](../../docs/mcp) +- **Team Spend Logs RBAC** — Teams with the `/spend/logs` permission can view team-wide spend logs from the UI and API +- **Bulk Team Permissions API** — New `POST /team/permissions_bulk_update` endpoint for updating member permissions across many teams in one call +- **Azure Container Routing** — Container routing, managed container IDs, and delete-response parsing for Azure Responses API containers +- **UI E2E Test Suite** — Playwright-based end-to-end tests for proxy admin, team, and key management flows now run in CI + +--- + +## New Models / Updated Models + +#### New Model Support (14 new models) + +| Provider | Model | Context Window | Input ($/1M tokens) | Output ($/1M tokens) | Features | +| -------- | ----- | -------------- | ------------------- | -------------------- | -------- | +| AWS Bedrock (GovCloud) | `bedrock/us-gov-east-1/anthropic.claude-sonnet-4-5-20250929-v1:0` | 200K | $3.30 | $16.50 | Chat, vision, tool use, prompt caching, reasoning | +| AWS Bedrock (GovCloud) | `bedrock/us-gov-west-1/anthropic.claude-sonnet-4-5-20250929-v1:0` | 200K | $3.30 | $16.50 | Chat, vision, tool use, prompt caching, reasoning | +| AWS Bedrock (GovCloud) | `us-gov.anthropic.claude-sonnet-4-5-20250929-v1:0` | 200K | $3.30 | $16.50 | Bedrock Converse, with above-200K tier pricing | +| Baseten | `baseten/MiniMaxAI/MiniMax-M2.5` | - | $0.30 | $1.20 | Chat | +| Baseten | `baseten/nvidia/Nemotron-120B-A12B` | - | $0.30 | $0.75 | Chat | +| Baseten | `baseten/zai-org/GLM-5` | - | $0.95 | $3.15 | Chat | +| Baseten | `baseten/zai-org/GLM-4.7` | - | $0.60 | $2.20 | Chat | +| Baseten | `baseten/zai-org/GLM-4.6` | - | $0.60 | $2.20 | Chat | +| Baseten | `baseten/moonshotai/Kimi-K2.5` | - | $0.60 | $3.00 | Chat | +| Baseten | `baseten/moonshotai/Kimi-K2-Thinking` | - | $0.60 | $2.50 | Chat | +| Baseten | `baseten/moonshotai/Kimi-K2-Instruct-0905` | - | $0.60 | $2.50 | Chat | +| Baseten | `baseten/openai/gpt-oss-120b` | - | $0.10 | $0.50 | Chat | +| Baseten | `baseten/deepseek-ai/DeepSeek-V3.1` | - | $0.50 | $1.50 | Chat | +| Baseten | `baseten/deepseek-ai/DeepSeek-V3-0324` | - | $0.77 | $0.77 | Chat | + +#### Features + +- **[AWS Bedrock](../../docs/providers/bedrock)** + - Update GovCloud Claude Sonnet 4.5 pricing, raise `max_tokens` to 8192, and add prompt-caching costs + - Skip dummy `user` continue message when assistant prefix prefill is set - [PR #25419](https://github.com/BerriAI/litellm/pull/25419) + - Avoid double-counting cache tokens in Anthropic Messages streaming usage - [PR #25517](https://github.com/BerriAI/litellm/pull/25517) +- **[Anthropic](../../docs/providers/anthropic)** + - Support `advisor_20260301` tool type - [PR #25525](https://github.com/BerriAI/litellm/pull/25525) +- **[Google Gemini / Vertex AI](../../docs/providers/gemini)** + - Mark applicable Gemini 2.5/3 models with `supports_service_tier` + +### Bug Fixes + +- **[AWS Bedrock](../../docs/providers/bedrock)** + - Pass-through fix for Bedrock JSON body and multipart uploads - [PR #25464](https://github.com/BerriAI/litellm/pull/25464) +- **[OpenAI](../../docs/providers/openai)** + - Mock headers in `test_completion_fine_tuned_model` to stabilize tests - [PR #25444](https://github.com/BerriAI/litellm/pull/25444) + +## LLM API Endpoints + +#### Features + +- **[Responses API](../../docs/response_api)** + - Containers: Azure routing, managed container IDs, and delete-response parsing - [PR #25287](https://github.com/BerriAI/litellm/pull/25287) + - WebSocket: append `?model=` to backend WebSocket URL so model selection routes correctly - [PR #25437](https://github.com/BerriAI/litellm/pull/25437) +- **[OpenAI / Files API](../../docs/providers/openai)** + - Add file content streaming support for OpenAI and related utilities - [PR #25450](https://github.com/BerriAI/litellm/pull/25450) +- **[A2A](../../docs/mcp)** + - Default 60-second timeout when creating an A2A client - [PR #25514](https://github.com/BerriAI/litellm/pull/25514) + +#### Bugs + +- **[Responses API](../../docs/response_api)** + - Map refusal `stop_reason` to `incomplete` status in streaming - [PR #25498](https://github.com/BerriAI/litellm/pull/25498) + - Fix duplicate keyword argument error in Responses WebSocket path - [PR #25513](https://github.com/BerriAI/litellm/pull/25513) +- **General** + - Ensure spend/cost logging runs when `stream=True` for web-search interception - [PR #25424](https://github.com/BerriAI/litellm/pull/25424) + +## Management Endpoints / UI + +#### Features + +- **Teams + Organizations** + - New `POST /team/permissions_bulk_update` endpoint for bulk permission updates across teams - [PR #25239](https://github.com/BerriAI/litellm/pull/25239) + - Team member permission `/spend/logs` to view team-wide spend logs (UI + RBAC) - [PR #25458](https://github.com/BerriAI/litellm/pull/25458) + - Align org and team endpoint permission checks - [PR #25554](https://github.com/BerriAI/litellm/pull/25554) +- **Virtual Keys** + - Align `/v2/key/info` response handling with v1 - [PR #25313](https://github.com/BerriAI/litellm/pull/25313) +- **Authentication / Routing** + - Consolidate route auth for UI and API tokens - [PR #25473](https://github.com/BerriAI/litellm/pull/25473) + - Use parameterized query for `combined_view` token lookup - [PR #25467](https://github.com/BerriAI/litellm/pull/25467) +- **Provider Credentials** + - Per-team / per-project credential overrides via `model_config` metadata - [PR #24438](https://github.com/BerriAI/litellm/pull/24438) +- **UI** + - Improve browser storage handling and Dockerfile consistency - [PR #25384](https://github.com/BerriAI/litellm/pull/25384) + - Align v1 guardrail and agent list responses with v2 field handling - [PR #25478](https://github.com/BerriAI/litellm/pull/25478) + - Flush Tremor Tooltip timers in `user_edit_view` tests - [PR #25480](https://github.com/BerriAI/litellm/pull/25480) + +#### Bugs + +- Improve input validation on management endpoints - [PR #25445](https://github.com/BerriAI/litellm/pull/25445) +- Harden file path resolution in skill archive extraction - [PR #25475](https://github.com/BerriAI/litellm/pull/25475) + +## AI Integrations + +### Logging + +- **[Langfuse](../../docs/proxy/logging#langfuse)** + - Preserve proxy key-auth metadata on `/v1/messages` Langfuse traces - [PR #25448](https://github.com/BerriAI/litellm/pull/25448) +- **[Prometheus](../../docs/proxy/logging#prometheus)** + - Reduce default `LATENCY_BUCKETS` from 35 → 18 boundaries (see breaking-change note above) - [PR #25527](https://github.com/BerriAI/litellm/pull/25527) +- **General** + - S3 logging: retry with exponential backoff for transient 503/500 errors - [PR #25530](https://github.com/BerriAI/litellm/pull/25530) + +### Guardrails + +- Optional skip system message in unified guardrail inputs - [PR #25481](https://github.com/BerriAI/litellm/pull/25481) +- Inline IAM: apply guardrail support - [PR #25241](https://github.com/BerriAI/litellm/pull/25241) +- Preserve `dict` `HTTPException.detail` and Bedrock context in guardrail errors - [PR #25558](https://github.com/BerriAI/litellm/pull/25558) + +## Spend Tracking, Budgets and Rate Limiting + +- Session-TZ-independent date filtering for spend / error log queries - [PR #25542](https://github.com/BerriAI/litellm/pull/25542) + +## MCP Gateway + +- **Per-user OAuth token storage for interactive MCP flows** - [PR #25441](https://github.com/BerriAI/litellm/pull/25441) +- Block arbitrary command execution via MCP `stdio` transport - [PR #25343](https://github.com/BerriAI/litellm/pull/25343) +- Document missing MCP per-user token environment variables in `config_settings` - [PR #25471](https://github.com/BerriAI/litellm/pull/25471) + +## Performance / Loadbalancing / Reliability improvements + +- Reduce Prometheus latency histogram cardinality (default buckets 35 → 18) - [PR #25527](https://github.com/BerriAI/litellm/pull/25527) +- S3 retry with exponential backoff for transient errors - [PR #25530](https://github.com/BerriAI/litellm/pull/25530) + +## Documentation Updates + +- Add Docker Image Security Guide covering cosign verification and deployment best practices - [PR #25439](https://github.com/BerriAI/litellm/pull/25439) +- Document April townhall announcements - [PR #25537](https://github.com/BerriAI/litellm/pull/25537) +- Document missing MCP per-user token env vars - [PR #25471](https://github.com/BerriAI/litellm/pull/25471) +- Add "Screenshots / Proof of Fix" section to PR template - [PR #25564](https://github.com/BerriAI/litellm/pull/25564) + +## Infrastructure / Security Notes + +- Pin cosign.pub verification to initial commit hash - [PR #25273](https://github.com/BerriAI/litellm/pull/25273) +- Fix node-gyp symlink path after npm upgrade in Dockerfile - [PR #25048](https://github.com/BerriAI/litellm/pull/25048) +- `Dockerfile.non_root`: handle missing `.npmrc` gracefully - [PR #25307](https://github.com/BerriAI/litellm/pull/25307) +- Add Playwright E2E tests with local PostgreSQL - [PR #25126](https://github.com/BerriAI/litellm/pull/25126) +- UI E2E tests for proxy admin team and key management - [PR #25365](https://github.com/BerriAI/litellm/pull/25365) +- Migrate Redis caching tests from GHA to CircleCI - [PR #25354](https://github.com/BerriAI/litellm/pull/25354) +- Update `check_responses_cost` tests for `_expire_stale_rows` - [PR #25299](https://github.com/BerriAI/litellm/pull/25299) +- Raise global vitest timeout and remove per-test overrides - [PR #25468](https://github.com/BerriAI/litellm/pull/25468) +- Version bumps and UI rebuilds: [PR #25316](https://github.com/BerriAI/litellm/pull/25316), [PR #25528](https://github.com/BerriAI/litellm/pull/25528), [PR #25578](https://github.com/BerriAI/litellm/pull/25578), [PR #25571](https://github.com/BerriAI/litellm/pull/25571), [PR #25573](https://github.com/BerriAI/litellm/pull/25573), [PR #25577](https://github.com/BerriAI/litellm/pull/25577) + +## New Contributors + +* @csoni-cweave made their first contribution in https://github.com/BerriAI/litellm/pull/25441 +* @jimmychen-p72 made their first contribution in https://github.com/BerriAI/litellm/pull/25530 + +**Full Changelog**: https://github.com/BerriAI/litellm/compare/v1.83.3.rc.1...v1.83.7.rc.1 From 4a1da629fac72363b6d1a76ceb943b55ca444d15 Mon Sep 17 00:00:00 2001 From: Yuneng Jiang Date: Tue, 14 Apr 2026 16:00:27 -0700 Subject: [PATCH 2/4] [Fix] Correct pip install versions for v1.83.3-stable and v1.83.7.rc.1 docs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit PyPI publishes 1.83.3 and 1.83.7 (no .post1 / rc1 suffixes) — align the pip install commands with the actual published versions. --- docs/my-website/release_notes/v1.83.3/index.md | 2 +- docs/my-website/release_notes/v1.83.7.rc.1/index.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/my-website/release_notes/v1.83.3/index.md b/docs/my-website/release_notes/v1.83.3/index.md index 1eced9239c9..9eead59b285 100644 --- a/docs/my-website/release_notes/v1.83.3/index.md +++ b/docs/my-website/release_notes/v1.83.3/index.md @@ -45,7 +45,7 @@ docker.litellm.ai/berriai/litellm:main-v1.83.3-stable ```bash -pip install litellm==1.83.3.post1 +pip install litellm==1.83.3 ``` diff --git a/docs/my-website/release_notes/v1.83.7.rc.1/index.md b/docs/my-website/release_notes/v1.83.7.rc.1/index.md index 811b129d22b..f9dcbf8c243 100644 --- a/docs/my-website/release_notes/v1.83.7.rc.1/index.md +++ b/docs/my-website/release_notes/v1.83.7.rc.1/index.md @@ -45,7 +45,7 @@ docker.litellm.ai/berriai/litellm:main-v1.83.7.rc.1 ```bash -pip install litellm==1.83.7rc1 +pip install litellm==1.83.7 ``` From 966be2982a568b8dd1477ca1d4bc2b35f5cbb4f2 Mon Sep 17 00:00:00 2001 From: Yuneng Jiang Date: Tue, 14 Apr 2026 16:13:09 -0700 Subject: [PATCH 3/4] [Docs] Add missed content PRs to v1.83.7.rc.1 and update runbook - Add 8 content PRs that merged directly to the release branch outside the listed staging PRs: #23769 (Ramp callback), #25252 (JWT OAuth2 override), #25254 (AWS GovCloud mode), #25258 (batch-limit cleanup), #25334 (router custom_llm_provider), #25345 (Triton embeddings), #25347 (tag-based routing), #25358 (Baseten pricing attribution) - Add @kedarthakkar to new contributors (first-ever PR via #23769) - Update RELEASE_NOTES_GENERATION_INSTRUCTIONS: require walking git log range between release tags in addition to staging PRs, and verify new-contributor status per author rather than trusting the GH release body floor --- .../RELEASE_NOTES_GENERATION_INSTRUCTIONS.md | 17 +++++++++++++++++ .../release_notes/v1.83.7.rc.1/index.md | 13 +++++++++++++ 2 files changed, 30 insertions(+) diff --git a/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md b/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md index ef7b146fe07..4a6fa9367fc 100644 --- a/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md +++ b/cookbook/misc/RELEASE_NOTES_GENERATION_INSTRUCTIONS.md @@ -18,6 +18,23 @@ The GitHub release page (e.g. `https://github.com/BerriAI/litellm/releases/tag/v To get the real changelog, you MUST click into each staging PR (e.g. `#23686`, `#23822`), open its **Commits** tab, and extract every underlying commit/PR (look for the `(#NNNNN)` suffix on commit titles). Those underlying PRs — not the staging PRs — are what get categorized in the release notes. Never treat a staging PR title as a single changelog entry. +**IMPORTANT — staging PRs are not the complete source.** Some PRs land on the release branch *before* the staging PRs and are therefore not reachable via `gh api /pulls//commits`. GitHub's auto-generated "What's Changed" on the release page also misses these. To catch every PR in the release, you MUST additionally walk the full git log range between the previous release's commit and this release's commit: + +```bash +git fetch origin --tags +git log .. --oneline | grep -oE '#[0-9]+' | sort -u +``` + +Union the PR set from the staging-PR walk with the PR set from `git log`. Any PR in `git log` but missing from your staging-expanded set is almost certainly a content PR that merged directly to the release branch — fetch its title/body with `gh pr view ` and categorize it. Do not trust the GH release body or the staging PRs alone as the authoritative list. + +**Sanity check for new contributors.** The GH release body's "New Contributors" list is a *floor*, not authoritative. For every PR author who appears in the release (including underlying PRs from staging and PRs found only via `git log`), verify whether they are a first-time contributor by running: + +```bash +gh api "search/issues?q=is:pr+author:+repo:BerriAI/litellm+is:merged&sort=created&order=asc" --jq '.items[0] | {n:.number, merged:.closed_at}' +``` + +If the author's earliest merged PR number matches a PR in this release window, they are a new contributor. If their earliest merged PR predates the previous release tag, they are not. Do not copy the GH release body's list blindly — it can both miss contributors (PRs that merged via an older dev branch) and falsely include contributors whose "first" PR in this window was not actually their first ever. + ## Step-by-Step Process ### 1. Initial Setup and Analysis diff --git a/docs/my-website/release_notes/v1.83.7.rc.1/index.md b/docs/my-website/release_notes/v1.83.7.rc.1/index.md index f9dcbf8c243..5fb41841498 100644 --- a/docs/my-website/release_notes/v1.83.7.rc.1/index.md +++ b/docs/my-website/release_notes/v1.83.7.rc.1/index.md @@ -91,11 +91,16 @@ pip install litellm==1.83.7 #### Features - **[AWS Bedrock](../../docs/providers/bedrock)** + - AWS GovCloud mode support (`us-gov` prefix routing) - [PR #25254](https://github.com/BerriAI/litellm/pull/25254) - Update GovCloud Claude Sonnet 4.5 pricing, raise `max_tokens` to 8192, and add prompt-caching costs - Skip dummy `user` continue message when assistant prefix prefill is set - [PR #25419](https://github.com/BerriAI/litellm/pull/25419) - Avoid double-counting cache tokens in Anthropic Messages streaming usage - [PR #25517](https://github.com/BerriAI/litellm/pull/25517) - **[Anthropic](../../docs/providers/anthropic)** - Support `advisor_20260301` tool type - [PR #25525](https://github.com/BerriAI/litellm/pull/25525) +- **[Triton](../../docs/providers/triton-inference-server)** + - Embedding usage estimation for self-hosted Triton responses - [PR #25345](https://github.com/BerriAI/litellm/pull/25345) +- **[Baseten](../../docs/providers/baseten)** + - Add pricing entries for 11 new Baseten-hosted models - [PR #25358](https://github.com/BerriAI/litellm/pull/25358) - **[Google Gemini / Vertex AI](../../docs/providers/gemini)** - Mark applicable Gemini 2.5/3 models with `supports_service_tier` @@ -123,6 +128,9 @@ pip install litellm==1.83.7 - **[Responses API](../../docs/response_api)** - Map refusal `stop_reason` to `incomplete` status in streaming - [PR #25498](https://github.com/BerriAI/litellm/pull/25498) - Fix duplicate keyword argument error in Responses WebSocket path - [PR #25513](https://github.com/BerriAI/litellm/pull/25513) +- **Router** + - Pass `custom_llm_provider` to `get_llm_provider` for unprefixed model names - [PR #25334](https://github.com/BerriAI/litellm/pull/25334) + - Fix tag-based routing when `encrypted_content_affinity` is enabled - [PR #25347](https://github.com/BerriAI/litellm/pull/25347) - **General** - Ensure spend/cost logging runs when `stream=True` for web-search interception - [PR #25424](https://github.com/BerriAI/litellm/pull/25424) @@ -137,6 +145,7 @@ pip install litellm==1.83.7 - **Virtual Keys** - Align `/v2/key/info` response handling with v1 - [PR #25313](https://github.com/BerriAI/litellm/pull/25313) - **Authentication / Routing** + - Allow JWT to override OAuth2 routing without requiring global OAuth2 enablement - [PR #25252](https://github.com/BerriAI/litellm/pull/25252) - Consolidate route auth for UI and API tokens - [PR #25473](https://github.com/BerriAI/litellm/pull/25473) - Use parameterized query for `combined_view` token lookup - [PR #25467](https://github.com/BerriAI/litellm/pull/25467) - **Provider Credentials** @@ -155,6 +164,8 @@ pip install litellm==1.83.7 ### Logging +- **[Ramp](../../docs/proxy/logging)** + - Add Ramp as a built-in success callback - [PR #23769](https://github.com/BerriAI/litellm/pull/23769) - **[Langfuse](../../docs/proxy/logging#langfuse)** - Preserve proxy key-auth metadata on `/v1/messages` Langfuse traces - [PR #25448](https://github.com/BerriAI/litellm/pull/25448) - **[Prometheus](../../docs/proxy/logging#prometheus)** @@ -171,6 +182,7 @@ pip install litellm==1.83.7 ## Spend Tracking, Budgets and Rate Limiting - Session-TZ-independent date filtering for spend / error log queries - [PR #25542](https://github.com/BerriAI/litellm/pull/25542) +- Batch-limit stale managed-object cleanup to prevent 300K+ row updates - [PR #25258](https://github.com/BerriAI/litellm/pull/25258) ## MCP Gateway @@ -204,6 +216,7 @@ pip install litellm==1.83.7 ## New Contributors +* @kedarthakkar made their first contribution in https://github.com/BerriAI/litellm/pull/23769 * @csoni-cweave made their first contribution in https://github.com/BerriAI/litellm/pull/25441 * @jimmychen-p72 made their first contribution in https://github.com/BerriAI/litellm/pull/25530 From 3aae15f5d829eb711988b44ad09866f7c789ec77 Mon Sep 17 00:00:00 2001 From: Ryan Crabbe Date: Tue, 14 Apr 2026 16:22:07 -0700 Subject: [PATCH 4/4] [Docs] Use GitHub avatar for Ryan Crabbe in release notes Replace the expiring LinkedIn CDN image URL with a stable GitHub avatar URL for v1.83.3 and v1.83.7.rc.1 release notes. --- docs/my-website/release_notes/v1.83.3/index.md | 2 +- docs/my-website/release_notes/v1.83.7.rc.1/index.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/my-website/release_notes/v1.83.3/index.md b/docs/my-website/release_notes/v1.83.3/index.md index 9eead59b285..c93648f9a92 100644 --- a/docs/my-website/release_notes/v1.83.3/index.md +++ b/docs/my-website/release_notes/v1.83.3/index.md @@ -14,7 +14,7 @@ authors: - name: Ryan Crabbe title: Full Stack Engineer, LiteLLM url: https://www.linkedin.com/in/ryan-crabbe-0b9687214 - image_url: https://media.licdn.com/dms/image/v2/D5603AQHt1t9Z4BJ6Gw/profile-displayphoto-shrink_400_400/profile-displayphoto-shrink_400_400/0/1724453682340?e=1772064000&v=beta&t=VXdmr13rsNB05wyA2F1TENOB5UuDHUZ0FCHTolNyR5M + image_url: https://github.com/ryan-crabbe.png - name: Yuneng Jiang title: Senior Full Stack Engineer, LiteLLM url: https://www.linkedin.com/in/yuneng-david-jiang-455676139/ diff --git a/docs/my-website/release_notes/v1.83.7.rc.1/index.md b/docs/my-website/release_notes/v1.83.7.rc.1/index.md index 5fb41841498..3b72e031b63 100644 --- a/docs/my-website/release_notes/v1.83.7.rc.1/index.md +++ b/docs/my-website/release_notes/v1.83.7.rc.1/index.md @@ -14,7 +14,7 @@ authors: - name: Ryan Crabbe title: Full Stack Engineer, LiteLLM url: https://www.linkedin.com/in/ryan-crabbe-0b9687214 - image_url: https://media.licdn.com/dms/image/v2/D5603AQHt1t9Z4BJ6Gw/profile-displayphoto-shrink_400_400/profile-displayphoto-shrink_400_400/0/1724453682340?e=1772064000&v=beta&t=VXdmr13rsNB05wyA2F1TENOB5UuDHUZ0FCHTolNyR5M + image_url: https://github.com/ryan-crabbe.png - name: Yuneng Jiang title: Senior Full Stack Engineer, LiteLLM url: https://www.linkedin.com/in/yuneng-david-jiang-455676139/