v1.83.0-nightly
<h2>Verifying the Docker Image</h2> <div class="snippet-clipboard-content notranslate position-relative overflow-auto" data-snippet-clipboard-copy-content="cosign verify --insecure-ignore-tlog=true --key https://raw.githubusercontent.com/BerriAI/litellm/main/cosign.pub \ ghcr.io/berriai/litellm:v1.83.0-nightly"><pre class="notranslate"><code>cosign verify --insecure-ignore-tlog=true --key https://raw.githubusercontent.com/BerriAI/litellm/main/cosign.pub \ ghcr.io/berriai/litellm:v1.83.0-nightly </code></pre></div> <p>Output:</p> <div class="snippet-clipboard-content notranslate position-relative overflow-auto" data-snippet-clipboard-copy-content="Verification for ghcr.io/berriai/litellm:v1.83.0-nightly -- The following checks were performed on each of these signatures: - The cosign claims
Verifying the Docker Image
cosign verify --insecure-ignore-tlog=true --key https://raw.githubusercontent.com/BerriAI/litellm/main/cosign.pub \ ghcr.io/berriai/litellm:v1.83.0-nightlycosign verify --insecure-ignore-tlog=true --key https://raw.githubusercontent.com/BerriAI/litellm/main/cosign.pub \ ghcr.io/berriai/litellm:v1.83.0-nightlyOutput:
Verification for ghcr.io/berriai/litellm:v1.83.0-nightly -- The following checks were performed on each of these signatures:Verification for ghcr.io/berriai/litellm:v1.83.0-nightly -- The following checks were performed on each of these signatures:- The cosign claims were validated
- The signatures were verified against the specified public key`
What's Changed
-
ci: skip scheduled workflows on forks by @DmitriyAlergant in #24460
-
fix(proxy): ignore return_to in SSO when control_plane_url is not con… by @ryan-crabbe in #24475
-
Correct documentation of completion_model by @benlangfeld in #24413
-
chore: migrate route_preview.tsx from Tremor to Ant Design by @ryan-crabbe in #24485
-
docs: add audit log export to S3 documentation by @ryan-crabbe in #24486
-
fix(security_scans.sh): pin trivvy version by @krrish-berri-2 in #24525
-
docs: add security update blog post for March 2026 supply chain incident by @ryan-crabbe-berri in #24537
-
docs: minor updates to security update blog post by @ishaan-berri in #24540
-
ci: remove publish-migrations and reset_stable workflows by @ishaan-berri in #24541
-
security: remove .claude/settings.json and block re-adding via semgrep by @ishaan-berri in #24584
-
docs: remove phone numbers from readme and docs by @ishaan-berri in #24587
-
Litellm ryan march 23 by @ryan-crabbe in #24490
-
[Infra] Pins GitHub Action workflows to specific hashes by @yuneng-berri in #24594
-
docs(opencode): add guidance for dropping reasoningSummary param by @Sameerlite in #24468
-
[Infra] Pin GHA dependencies and remove unused load test files by @yuneng-berri in #24607
-
Update README.md by @ishaan-berri in #24649
-
[Infra] Migrate PyPI Publishing from CircleCI to GitHub Actions OIDC by @yuneng-berri in #24654
-
Add zizmor to ci/cd by @krrish-berri-2 in #24663
-
feat(prometheus): add metrics for managed batch lifecycle by @Sameerlite in #24691
-
feat(openai): round-trip Responses API reasoning_items in chat completions by @Sameerlite in #24690
-
feat(fine-tuning): fix Azure OpenAI fine-tuning job creation by @Sameerlite in #24687
-
feat(gemini): add gemini-3.1-flash-live-preview to model cost map by @Sameerlite in #24665
-
feat(gemini): normalize AI Studio file retrieve URL by @Sameerlite in #24662
-
fix(anthropic): strip undocumented keys from metadata before sending to API by @Sameerlite in #24661
-
Litellm security townhall blog by @krrish-berri-2 in #24692
-
fix(proxy): sanitize user_id input and block dangerous env var keys by @Sameerlite in #24624
-
feat(gemini): Lyria 3 preview models in cost map and docs by @Sameerlite in #24610
-
fix(openrouter): strip routing prefix for wildcard proxy deployments by @Sameerlite in #24603
-
fix(pricing): remove above_200k_tokens price tiers for claude-opus-4-6 and claude-sonnet-4-6 by @Sameerlite in #24689
-
fix(team-routing): preserve sibling deployment candidates for team public models by @Sameerlite in #24688
-
[Fix] Pin Prisma Node.js dependency in CI workflows by @yuneng-berri in #24696
-
[Infra] Improve CodeQL scanning coverage and schedule by @yuneng-berri in #24697
-
fix(proxy): enforce budget limits across multi-pod deployments via Redis-backed spend counters by @michelligabriele in #24682
-
[Infra] Automated schema.prisma sync and drift detection by @yuneng-berri in #24705
-
fix: add /user/bulk_update to management routes by @ryan-crabbe-berri in #24708
-
fix(sso): pass decoded JWT access token to role mapping during SSO login by @ryan-crabbe-berri in #24701
-
fix(auth): guard JWTHandler.is_jwt() against None token by @ryan-crabbe-berri in #24706
-
fix(ui): refactor budget page to React Query hooks and fix crashes by @ryan-crabbe-berri in #24711
-
fix(jwt): invalidate user cache after role/team sync updates by @ryan-crabbe-berri in #24717
-
litellm ryan march 26 by @ryan-crabbe-berri in #24718
-
[Infra] Isolate unit test workflows with hardened security posture by @yuneng-berri in #24740
-
[Fix] Test Isolation and Path Resolution for GHA Unit Tests by @yuneng-berri in #24741
-
[Infra] Add unit test workflows for Postgres, Redis, and security suites by @yuneng-berri in #24742
-
Feat/prometheus org budget metrics by @J-Byron in #24449
-
Bedrock: move native structured output model list to cost JSON, add Sonnet 4.6 by @ndgigliotti in #23794
-
[Infra] Remove CircleCI jobs now covered by GitHub Actions by @yuneng-berri in #24754
-
Litellm test cleanup by @krrish-berri-2 in #24755
-
[Fix] Remove NLP_CLOUD_API_KEY requirement from test_exceptions by @yuneng-berri in #24756
-
Litellm vanta announcement by @krrish-berri-2 in #24800
-
adopt OpenSSF Scorecard (https://scorecard.dev/) by @krrish-berri-2 in #24792
-
[Test] Mock DeepInfra completion tests to avoid real API calls by @yuneng-berri in #24805
-
[Refactor] Extract helper methods in guardrail handlers to fix PLR0915 by @yuneng-berri in #24802
-
[Fix] Mypy Type Errors in Responses Transformation, Spend Tracking, and PagerDuty by @yuneng-berri in #24803
-
[Fix] Update gemini-2.0-flash to gemini-2.5-flash in test_gemini by @yuneng-berri in #24817
-
[Fix] Router code coverage CI failure for health check filter tests by @yuneng-berri in #24812
-
[Fix] WatsonX Tests Failing on CI Due to Missing Env Vars by @yuneng-berri in #24814
-
[Fix] Correct kwarg name in test_user_api_key_auth tests by @yuneng-berri in #24820
-
fix(responses): emit content_part.added event for non-OpenAI models by @nielei3 in #24445
-
Fix returned model when batch completions is used - return picked model, not comma-separated list by @krrish-berri-2 in #24753
-
[Test] Move Snowflake Mocked Tests to Unit Test Directory by @yuneng-berri in #24822
-
[Fix] Use correct Redis env vars and fix Azure AD token test mocking by @yuneng-berri in #24824
-
chore: fixes by @jaydns in #24823
-
test: use dynamic db by @krrish-berri-2 in #24827
-
[Fix] Failure callbacks silently skipped when customLogger not initialized by @yuneng-berri in #24826
-
bump: litellm-proxy-extras 0.4.61 → 0.4.62 for schema changes by @krrish-berri-2 in #24828
-
[Security] - Pin axios=1.13.6 by @ishaan-berri in #24829
-
[Test] Move test_add_and_delete_model to Mock Test by @yuneng-berri in #24837
-
[Infra] Bump Version to 1.83.0 by @yuneng-berri in #24840
New Contributors
-
@benlangfeld made their first contribution in #24413
-
@J-Byron made their first contribution in #24449
-
@jaydns made their first contribution in #24823
Full Changelog: v1.82.6.dev2...v1.83.0-nightly
Sign in to highlight and annotate this article

Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
claudegeminimodelInside the hours when coders tore through Claude's guts and found pets, spinner verbs, and a curse chart - Business Insider
<a href="https://news.google.com/rss/articles/CBMipwFBVV95cUxNMl9tQk1icUo1Tzc4YXU2NmpOMjFUcEZtdEhNLVdSZ0kxaUtTRUtmbXdLOXV0MDJiSWFHTk5YYzhCS2dxT2dxTzFPdzBlZVU3MTBTcl94RnhPT1liNnExbzJXYVo1Y0NWRllfWUI1RkdENi00d09CMmNhU013WWhnWnNrSXQ0Y3FFMVRTdFpCcEdZby1XWEJmWXUxZWR5S295ekVPTGxscw?oc=5" target="_blank">Inside the hours when coders tore through Claude's guts and found pets, spinner verbs, and a curse chart</a> <font color="#6f6f6f">Business Insider</font>
A Very Fine Untuning
How fine-tuning made my chatbot worse (and broke my RAG pipeline) I spent weeks trying to improve my personal chatbot, Virtual Alexandra , with fine-tuning. Instead I got increased hallucination rate and broken retrieval in my RAG system. Yes, this is a story about a failed attempt, not a successful one. My husband and I called fine tuning results “Drunk Alexandra” — incoherent answers that were initially funny, but quickly became annoying. After weeks of experiments, I reached a simple conclusion: for this particular project, a small chatbot that answers questions based on my writing and instructions, fine tuning was not a good option. It was not just unnecessary, it actively degraded the experience and didn’t justify the extra time, cost, or complexity compared to the prompt + RAG system

Google's TurboQuant saves memory, but won't save us from DRAM-pricing hell
<h4>Chocolate Factory’s compression tech clears the way to cheaper AI inference, not more affordable memory</h4> <p>When Google unveiled <a target="_blank" rel="nofollow" href="https://research.google/blog/turboquant-redefining-ai-efficiency-with-extreme-compression/">TurboQuant</a>, an AI data compression technology that promises to slash the amount of memory required to serve models, many hoped it would help with a memory shortage that has seen prices triple since last year. Not so much.…</p>
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Models
Inside the hours when coders tore through Claude's guts and found pets, spinner verbs, and a curse chart - Business Insider
<a href="https://news.google.com/rss/articles/CBMipwFBVV95cUxNMl9tQk1icUo1Tzc4YXU2NmpOMjFUcEZtdEhNLVdSZ0kxaUtTRUtmbXdLOXV0MDJiSWFHTk5YYzhCS2dxT2dxTzFPdzBlZVU3MTBTcl94RnhPT1liNnExbzJXYVo1Y0NWRllfWUI1RkdENi00d09CMmNhU013WWhnWnNrSXQ0Y3FFMVRTdFpCcEdZby1XWEJmWXUxZWR5S295ekVPTGxscw?oc=5" target="_blank">Inside the hours when coders tore through Claude's guts and found pets, spinner verbs, and a curse chart</a> <font color="#6f6f6f">Business Insider</font>
Exclusive | The Sudden Fall of OpenAI’s Most Hyped Product Since ChatGPT - wsj.com
<a href="https://news.google.com/rss/articles/CBMiogNBVV95cUxQNU1NTGdtSzlqckVfLU5oMkoxamNrWjNya0oxOFNFQ3Q1aEdnSUlDN0lYY3ZKNEc5U0QtMmRXbFVyZEdpTHBwcllmbVFPZmR2WmlBSzh1TmJUN0tncFIzdXNmR2ZLNUtjc1NtYjlSNHJfMzI4NUFpQ3ZIVnJSRlJvVmZpa2U4dHJOTnVlMWVUZE9lSUNqN2ZuT3FzeE15TTlzZDIwWjkxVFhIS0hua2JDQm5pRjVBcWZhVjV5bkZIR2YxcmdkczFxMEJCcTEwQ2pQS2dhakVjdjRwOXhkbmFZV2dEM3dqUllySHJ6LXZtR21PNnRUQWxBVE11MjZ6ZkRmczNjbjAzLWlhZkFDZEJ3dkRiMnhybFhhYlluQVYtNUswX096NFNlOVptZzQ0VlB4bmx2a1ZQV0M0VE5sVDNKMWQtV1BlUzFxNENBaWYxNmlpOHdpbjVvWHZnZ2JVWndwbUFwbGRNSXhCRHFxMG53c09LZ3JkLUREb1FRLV8wcGptei0xemlKSDd4aU1oWnlkRUlwbzNaY196dmtoa1BIVlF3?oc=5" target="_blank">Exclusive | The Sudden Fall of OpenAI’s Most Hyped Product Since ChatGPT</a> <font color="#6f6f6f">wsj.com</font>
Exclusive | The Sudden Fall of OpenAI’s Most Hyped Product Since ChatGPT - WSJ
<a href="https://news.google.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?oc=5" target="_blank">Exclusive | The Sudden Fall of OpenAI’s Most Hyped Product Since ChatGPT</a> <font color="#6f6f6f">WSJ</font>
Android Auto left behind as ChatGPT comes to CarPlay - Android Authority
<a href="https://news.google.com/rss/articles/CBMihgFBVV95cUxQYzVSM3loaldFMW5RWlp3Qi1FenJLYkJJT0lBYVBYN1JyQlgxSHozbnJwVzBhZ1MyeDkyLUR0b2k1Y211ZFRwcC1ubW5fOXcyakhXRDU4OF9TMDVjdG00VDJOOUFDVzRiN21yZ2NBa1UwdEJqTE4yLS03anV0TEdjNVp3NHprdw?oc=5" target="_blank">Android Auto left behind as ChatGPT comes to CarPlay</a> <font color="#6f6f6f">Android Authority</font>
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!