v0.3.4

What's Changed

• chore: bump starlette version (backport #4158) by @mergify[bot] in #4248
• fix: uninitialised enable_write_queue by @leseb in #4264
• fix: Add policies to adapters (backport #4277) by @mergify[bot] in #4279
• fix: Avoid model_limits KeyError (backport #4060) by @mergify[bot] in #4283
• chore: bump mcp package version (backport #4287) by @mergify[bot] in #4288
• fix: RBAC bypass vulnerabilities in model access (backport #4270) by @mergify[bot] in #4285

Full Changelog: v0.3.3...v0.3.4

v0.3.3

What's Changed

• fix: allowed_models config did not filter models (backport #4030) by @mergify[bot] in #4223
• fix: Vector store persistence across server restarts (backport #3977) by @mergify[bot] in #4225
• fix: enable SQLite WAL mode to prevent database locking errors (backport #4048) by @mergify[bot] in #4226
• fix(docs): fix glob vulnerability (backport #4193) by @mergify[bot] in #4227
• fix: enforce allowed_models during inference requests (backport #4197) by @mergify[bot] in #4228
• fix: update hard-coded google model names (backport #4212) by @mergify[bot] in #4229

Full Changelog: v0.3.2...v0.3.3

v0.3.2

What's Changed

• fix: only set UV_INDEX_STRATEGY when UV_EXTRA_INDEX_URL is present by @ashwinb in #4017
• fix(ci): export UV_INDEX_STRATEGY to current shell before running uv sync by @ashwinb in #4019
• fix: print help for list-deps if no args (backport #4078) by @mergify[bot] in #4083
• docs: use 'uv pip' to avoid pitfalls of using 'pip' in virtual environment (backport #4122) by @mergify[bot] in #4136
• docs: clarify model identification uses provider_model_id not model_id (backport #4128) by @mergify[bot] in #4137
• chore(ci): remove unused recordings (backport #4074) by @mergify[bot] in #4141
• fix: harden storage semantics (backport #4118) by @mergify[bot] in #4138
• fix(inference): enable routing of models with provider_data alone (backport #3928) by @mergify[bot] in #4142

Full Changelog: v0.3.1...v0.3.2

v0.3.1

What's Changed

• feat(cherry-pick): fixes for 0.3.1 release by @ashwinb in #3998
• fix(ci): install client from release branch before uv sync by @ashwinb in #4002
• chore(release-0.3.x): handle missing external_providers_dir by @ashwinb in #4011
• fix(ci): unset empty UV index env vars to prevent uv errors by @ashwinb in #4013
• feat: support workers in run config by @ashwinb in #4014
• docs: A getting started notebook featuring simple agent examples by @ashwinb in #4015

Full Changelog: v0.3.0...v0.3.1

v0.3.0

Highlights

• Stable OpenAI-Compatible APIs
• Llama Stack now separates APIs into stable (/v1/), experimental (/v1alpha/ and /v1beta/) and deprecated (deprecated = True.)
• extra_body/metadata support for APIs which support extra functionality compared to the OpenAI implementation
• Documentation overhaul: Migration to Docusaurus, modern formatting, and improved API docs

What's Changed

• feat(internal): add image_url download feature to OpenAIMixin by @mattf in #3516
• chore(api): remove batch inference by @mattf in #3261
• chore(apis): unpublish deprecated /v1/inference apis by @mattf in #3297
• chore: recordings for fireworks (inference + openai) by @mattf in #3573
• chore: remove extra logging by @ehhuang in #3574
• chore: MANIFEST maintenance by @leseb in #3454
• feat: Add items and title to ToolParameter/ToolParamDefinition by @TamiTakamiya in #3003
• feat(ci): use @next branch from llama-stack-client by @ashwinb in #3576
• chore(ui-deps): bump shiki from 1.29.2 to 3.13.0 in /llama_stack/ui by @dependabot[bot] in #3585
• chore(ui-deps): bump tw-animate-css from 1.2.9 to 1.4.0 in /llama_stack/ui by @dependabot[bot] in #3583
• chore(github-deps): bump actions/cache from 4.2.4 to 4.3.0 by @dependabot[bot] in #3577
• chore: skip nvidia datastore tests when nvidia datastore is not enabled by @mattf in #3590
• chore: introduce write queue for response_store by @ehhuang in #3497
• revert: feat(ci): use @next branch from llama-stack-client by @ashwinb in #3593
• fix: adding mime type of application/json support by @wukaixingxp in #3452
• chore(api): remove deprecated embeddings impls by @mattf in #3301
• feat(api): level inference/rerank and remove experimental by @cdoern in #3565
• chore: skip safety tests when shield not available by @mattf in #3592
• feat: update eval runner to use openai endpoints by @mattf in #3588
• docs: update image paths by @reluctantfuturist in #3599
• fix: remove inference.completion from docs by @mattf in #3589
• fix: Remove deprecated user param in OpenAIResponseObject by @slekkala1 in #3596
• fix: ensure usage is requested if telemetry is enabled by @mhdawson in #3571
• feat(openai_movement): Change URL structures to kill /openai/v1 (part 1) by @ashwinb in #3587
• feat(files): fix expires_after API shape by @ashwinb in #3604
• feat(openai_movement)!: Change URL structures to kill /openai/v1 (part 2) by @ashwinb in #3605
• fix: mcp tool with array type should include items by @ehhuang in #3602
• feat: add llamastack + CrewAI integration example notebook by @wukaixingxp in #3275
• chore: unpublish /inference/chat-completion by @mattf in #3609
• feat: use /v1/chat/completions for safety model inference by @mattf in #3591
• feat(api): level /agents as v1alpha by @cdoern in #3610
• feat(api): Add Vector Store File batches api stub by @slekkala1 in #3615
• fix(expires_after): make sure multipart/form-data is properly parsed by @ashwinb in #3612
• docs: frontpage update by @reluctantfuturist in #3620
• docs: update safety notebook by @reluctantfuturist in #3617
• feat: add support for require_approval argument when creating response by @grs in #3608
• fix: don't pass default response format in Responses by @ehhuang in #3614
• fix(logging): disable console telemetry sink by default by @ashwinb in #3623
• fix: Ensure that tool calls with no arguments get handled correctly by @jaideepr97 in #3560
• chore: use openai_chat_completion for llm as a judge scoring by @mattf in #3635
• chore: remove /v1/inference/completion and implementations by @mattf in #3622
• feat(api): implement v1beta leveling, and additional alpha by @cdoern in #3594
• feat(conformance): skip test if breaking change is ack by @cdoern in #3619
• fix: log level by @ehhuang in #3637
• docs: update API conformance test by @reluctantfuturist in #3631
• docs: api separation by @reluctantfuturist in #3630
• docs: adding supplementary markdown content to API specs by @reluctantfuturist in #3632
• chore: add provider-data-api-key support to openaimixin by @mattf in #3639
• chore: Remove debug logging from telemetry adapter by @ehhuang in #3643
• docs: fix broken links by @reluctantfuturist in #3647
• docs: add favicon and mobile styling by @reluctantfuturist in #3650
• docs: fix more broken links by @reluctantfuturist in #3649
• docs: Fix Dell distro documentation code snippets by @ConnorHack in #3640
• refactor(agents): migrate to OpenAI chat completions API by @aakankshaduggal in #3323
• fix: re-enable conformance skipping ability by @cdoern in #3651
• chore!: add double routes for v1/openai/v1 by @leseb in #3636
• docs: Update docs navbar config by @kelbrown20 in #3653
• docs: API spec generation for Stainless by @reluctantfuturist in #3655
• chore: fix agents tests for non-ollama providers, provide max_tokens by @mattf in #3657
• chore: fix/add logging categories by @ehhuang in #3658
• chore: fix precommit by @ehhuang in #3663
• feat(tools)!: substantial clean up of "Tool" related datatypes by @ashwinb in #3627
• fix: responses <> chat completion input conversion by @ehhuang in #3645
• chore: OpenAIMixin implements ModelsProtocolPrivate by @mattf in #3662
• feat: auto-detect Console width by @rhdedgar in #3327
• feat: implement keyword and hybrid search for Weaviate provider by @ChristianZaccaria in #3264
• fix(docs): Correct indentation in documented example for access_policy by @anastasds in #3652
• chore: remove deprecated inference.chat_completion implementations by @mattf in #3654
• docs!: adjust external provider docs by @cdoern in #3484
• feat: Add OpenAI Conversations API by @franciscojavierarceo in #3429
• chore: use remoteinferenceproviderconfig for remote inference providers by @mattf in #3668
• docs: update OG image by @reluctantfuturist in #3669
• feat: add comment-triggered pre-commit bot for PRs by @ashwinb in #3672
• feat(api): add extra_body parameter support with shields example by @ashwinb in #3670
• chore: Add weaviate client to unit group in pyproject.toml and uv.lock by @franciscojavierarceo in #3675
• chore: update CODEOWNERS by @reluctantfuturist in #3613
• chore(tests): normalize recording IDs and timestamps to reduce git diff noise by @ashwinb in #3676
• chore: fix setup_telemetry script by @ehhuang in #3680
• docs: Update links in README for quick start and documentation by @seyeon...

v0.2.23

Highlights

• Overhauls documentation with Docusaurus migration and modern formatting.
• Standardizes Ollama and Fireworks provider with OpenAI compatibility layer.
• Combines dynamic model discovery with static embedding metadata for better model information.
• Refactors server.main for better code organization.
• Introduces API leveling with post_training and eval promoted to v1alpha.

What's Changed

• fix: Added a bug fix when registering new models by @omaryashraf5 in #3453
• fix: unbound variable PR_HEAD_REPO by @derekhiggins in #3469
• fix: Fixing prompts import warning by @franciscojavierarceo in #3455
• docs: update documentation links by @reluctantfuturist in #3459
• fix: Set provider_id in NVIDIA notebook when registering dataset by @JashG in #3472
• feat: update qdrant hash function from SHA-1 to SHA-256 by @rhdedgar in #3477
• feat: Add dynamic authentication token forwarding support for vLLM by @akram in #3388
• feat: include all models from provider's /v1/models by @mattf in #3471
• chore: update the ollama inference impl to use OpenAIMixin for openai-compat functions by @mattf in #3395
• fix: add missing files provider to NVIDIA distribution by @jiayin-nvidia in #3479
• feat: combine ProviderSpec datatypes by @cdoern in #3378
• chore: refactor server.main by @ehhuang in #3462
• docs: Fix incorrect vector_db_id usage in RAG tutorial by @adam-d-young in #3444
• fix: force milvus-lite installation for inline::milvus by @leseb in #3488
• chore: simplify authorized sqlstore by @ehhuang in #3496
• chore: remove duplicate AnthropicProviderDataValidator by @mattf in #3512
• fix: Update inference recorder to handle both Ollama and OpenAI model by @derekhiggins in #3470
• fix: handle missing API keys gracefully in model refresh by @derekhiggins in #3493
• chore: remove duplicate OpenAI and Gemini data validators by @mattf in #3513
• chore(github-deps): bump astral-sh/setup-uv from 6.6.1 to 6.7.0 by @dependabot[bot] in #3502
• chore(ui-deps): bump remeda from 2.30.0 to 2.32.0 in /llama_stack/ui by @dependabot[bot] in #3511
• chore(ui-deps): bump @radix-ui/react-dialog from 1.1.13 to 1.1.15 in /llama_stack/ui by @dependabot[bot] in #3510
• chore(ui-deps): bump jest-environment-jsdom from 29.7.0 to 30.1.2 in /llama_stack/ui by @dependabot[bot] in #3509
• fix: change ModelRegistryHelper to use ProviderModelEntry instead of hardcoded ModelType.llm by @wukaixingxp in #3451
• chore: Refactor fireworks to use OpenAIMixin by @slekkala1 in #3480
• chore: fix build by @ehhuang in #3522
• fix: return llama stack model id from embeddings by @mattf in #3525
• fix(dev): fix vllm inference recording (await models.list) by @mattf in #3524
• chore: refactor tracingmiddelware by @ehhuang in #3520
• feat: (re-)enable Databricks inference adapter by @mattf in #3500
• feat: update Cerebras inference provider to support dynamic model listing by @mattf in #3481
• docs: fix typos in RAG docs by @nathan-weinberg in #3530
• chore(perf): run guidellm benchmarks by @ehhuang in #3421
• fix: fix API docstrings for proper MDX parsing by @reluctantfuturist in #3526
• fix: update OpenAPI generator by @reluctantfuturist in #3527
• fix: update API conformance test to point to new schema location by @reluctantfuturist in #3528
• docs: provider and distro codegen migration by @reluctantfuturist in #3531
• docs: advanced_apis migration by @reluctantfuturist in #3532
• docs: concepts and building_applications migration by @reluctantfuturist in #3534
• docs: static content migration by @reluctantfuturist in #3535
• docs: docusaurus setup by @reluctantfuturist in #3541
• docs: MDX leftover fixes by @reluctantfuturist in #3536
• docs: fix broken links by @reluctantfuturist in #3540
• chore: use ollama/all-minilm:l6-v2 for ollama tests by @mattf in #3537
• feat: add static embedding metadata to dynamic model listings for providers using OpenAIMixin by @mattf in #3547
• fix: docs deployment URL by @reluctantfuturist in #3556
• feat: use FIPS validated CSPRNG for telemetry by @rhdedgar in #3554
• feat: together now supports base64 embedding encoding by @mattf in #3559
• feat: introduce API leveling, post_training, eval to v1alpha by @cdoern in #3449
• chore: prune mypy exclude list by @mattf in #3561
• fix: Gracefully handle errors when listing MCP tools by @bbrowning in #2544
• fix: ensure ModelRegistryHelper init for together and fireworks by @mattf in #3572
• fix: Revert "fix: Added a bug fix when registering new models" by @mattf in #3473

New Contributors

• @omaryashraf5 made their first contribution in #3453
• @adam-d-young made their first contribution in #3444

Full Changelog: v0.2.22...v0.2.23

v0.2.22

Highlights

• Migrated to unified "setups" system for test config
• Added default inference store automatically during llama stack build
• Introduced write queue for inference store
• Proposed API leveling framework
• Enhanced Together provider with embedding and dynamic model support

What's Changed

• feat(tests): migrate to global "setups" system for test configuration by @ashwinb in #3390
• chore: remove unused variable by @ehhuang in #3389
• feat: include a default inference store during llama stack build by @mattf in #3373
• feat: Add vector_db_id to chunk metadata by @are-ces in #3304
• fix: Add missing files_api parameter to MemoryToolRuntimeImpl test by @akram in #3394
• fix: pre-commit issues: non executable shebang file and removal of @pytest.mark.asyncio decorator by @akram in #3397
• chore: update the vertexai inference impl to use openai-python for openai-compat functions by @mattf in #3377
• ci: Re-enable pre-commit to fail by @leseb in #3399
• fix: Fireworks chat completion broken due to telemetry by @slekkala1 in #3392
• chore: logging perf improvments by @ehhuang in #3393
• revert: Fireworks chat completion broken due to telemetry by @franciscojavierarceo in #3402
• fix: unbound variable error in schedule-record-workflow.sh by @derekhiggins in #3401
• chore: introduce write queue for inference_store by @ehhuang in #3383
• docs: horizontal nav bar by @reluctantfuturist in #3407
• chore(python-deps): bump pytest from 8.4.1 to 8.4.2 by @dependabot[bot] in #3359
• chore(python-deps): bump locust from 2.39.1 to 2.40.1 by @dependabot[bot] in #3358
• chore(python-deps): bump openai from 1.102.0 to 1.106.1 by @dependabot[bot] in #3356
• chore(ui-deps): bump tailwindcss from 4.1.6 to 4.1.13 in /llama_stack/ui by @dependabot[bot] in #3362
• chore: telemetry test by @ehhuang in #3405
• chore: move benchmarking related code by @ehhuang in #3406
• fix(inference_store): on duplicate chat completion IDs, replace by @ashwinb in #3408
• chore: remove openai dependency from providers by @leseb in #3398
• fix: AWS Bedrock inference profile ID conversion for region-specific endpoints by @skamenan7 in #3386
• chore(replay): improve replay robustness with un-validated construction by @mattf in #3414
• feat: add Azure OpenAI inference provider support by @leseb in #3396
• chore: Updating documentation, adding exception handling for Vector Stores in RAG Tool, more tests on migration, and migrate off of inference_api for context_retriever for RAG by @franciscojavierarceo in #3367
• chore: update the vLLM inference impl to use OpenAIMixin for openai-compat functions by @mattf in #3404
• chore(unit tests): remove network use, update async test by @mattf in #3418
• feat: Add langchain llamastack Integration example notebook by @slekkala1 in #3314
• fix: oasdiff enhancements and stability by @cdoern in #3419
• fix: Improve pre-commit workflow error handling and feedback by @akram in #3400
• feat: migrate to FIPS-validated cryptographic algorithms by @rhdedgar in #3423
• chore(recorder, tests): add test for openai /v1/models by @mattf in #3426
• chore(tests): always show slowest tests by @mattf in #3431
• chore(recorder): add support for NOT_GIVEN by @mattf in #3430
• chore(ui-deps): bump next from 15.3.3 to 15.5.3 in /llama_stack/ui by @dependabot[bot] in #3438
• chore(ui-deps): bump @radix-ui/react-select from 2.2.5 to 2.2.6 in /llama_stack/ui by @dependabot[bot] in #3437
• chore(recorder): update mocks to be closer to non-mock environment by @mattf in #3442
• feat: create HTTP DELETE API endpoints to unregister ScoringFn and Benchmark resources in Llama Stack by @r3v5 in #3371
• feat: add dynamic model registration support to TGI inference by @mattf in #3417
• chore: various watsonx fixes by @leseb in #3428
• feat: introduce api leveling proposal by @cdoern in #3317
• fix: docker failing to start container[pydantic] by @slekkala1 in #3460
• feat: add embedding and dynamic model support to Together inference adapter by @mattf in #3458

New Contributors

• @rhdedgar made their first contribution in #3423

Full Changelog: v0.2.21...v0.2.22

v0.2.21

Highlights

• Testing infrastructure improvements and fixes
• Backwards compatibility tests for core APIs
• Added OpenAI Prompts API
• Updated RAG Tool to use Files API and Vector Stores API
• Descriptive MCP server connection errors

What's Changed

• feat(files, s3, expiration): add expires_after support to S3 files provider by @mattf in #3283
• docs: add VLM NIM example by @jiayin-nvidia in #3277
• chore(migrate apis): move VectorDBWithIndex from embeddings to openai_embeddings by @mattf in #3294
• chore(ui-deps): bump framer-motion from 11.18.2 to 12.23.12 in /llama_stack/ui by @dependabot[bot] in #3291
• chore(ui-deps): bump @types/node from 20.17.47 to 24.3.0 in /llama_stack/ui by @dependabot[bot] in #3290
• chore(ui-deps): bump eslint-config-next from 15.3.2 to 15.5.2 in /llama_stack/ui by @dependabot[bot] in #3288
• chore(ui-deps): bump prettier from 3.5.3 to 3.6.2 in /llama_stack/ui by @dependabot[bot] in #3289
• chore(ui-deps): bump @radix-ui/react-tooltip from 1.2.6 to 1.2.8 in /llama_stack/ui by @dependabot[bot] in #3287
• chore(python-deps): bump locust from 2.39.0 to 2.39.1 by @dependabot[bot] in #3284
• refactor: remove lama-api-client from pyproject.toml by @r3v5 in #3299
• chore(python-deps): bump pymilvus from 2.6.0 to 2.6.1 by @dependabot[bot] in #3285
• refactor: use generic WeightedInMemoryAggregator for hybrid search in SQLiteVecIndex by @r3v5 in #3303
• fix: Fix mock vector DB schema in Qdrant tests by @varshaprasad96 in #3295
• chore(python-deps): replace ibm_watson_machine_learning with ibm_watsonx_ai by @are-ces in #3302
• chore: Improve error message for missing provider dependencies by @ehhuang in #3315
• feat(tests): auto-merge all model list responses and unify recordings by @ashwinb in #3320
• fix(tests): set inference mode to be replay by default by @ashwinb in #3326
• chore: handle missing finish_reason by @ehhuang in #3328
• fix: distro-codegen pre-commit hook file pattern by @derekhiggins in #3337
• refactor(server): remove hardcoded 409 and 404 status codes in server.py using httpx constants by @r3v5 in #3333
• fix: Make SentenceTransformer embedding operations non-blocking by @derekhiggins in #3335
• chore: async inference store write by @ehhuang in #3318
• fix: Move to older version for docker container failure [fireworks-ai] by @slekkala1 in #3338
• fix: show descriptive MCP server connection errors instead of generic 500s by @skamenan7 in #3256
• chore: unbreak inference store test by @ehhuang in #3340
• fix: use lambda pattern for bedrock config env vars by @skamenan7 in #3307
• fix: Fix locations of distrubution runtime directories by @derekhiggins in #3336
• feat!: Migrate Vector DB IDs to Vector Store IDs (breaking change) by @franciscojavierarceo in #3253
• feat(batches, completions): add /v1/completions support to /v1/batches by @mattf in #3309
• chore(sambanova test): skip with_n tests for sambanova, it is not implemented server-side by @mattf in #3342
• feat(tests): introduce a test "suite" concept to encompass dirs, options by @ashwinb in #3339
• feat: Updating Rag Tool to use Files API and Vector Stores API by @franciscojavierarceo in #3344
• chore: update the gemini inference impl to use openai-python for openai-compat functions by @mattf in #3351
• chore(gemini, tests): add skips for n and completions, gemini api does not support them by @mattf in #3350
• chore: update the sambanova inference impl to use openai-python for openai-compat functions by @mattf in #3345
• chore(groq test): skip with_n tests for groq, it is not supported server-side by @mattf in #3346
• test: introduce api conformance test by @cdoern in #3257
• chore: update the groq inference impl to use openai-python for openai-compat functions by @mattf in #3348
• chore(groq test): skip completions tests for groq, api is not supported server-side by @mattf in #3347
• chore: update the anthropic inference impl to use openai-python for openai-compat functions by @mattf in #3366
• chore(ui-deps): bump react-dom and @types/react-dom in /llama_stack/ui by @dependabot[bot] in #3360
• chore(ui-deps): bump sonner from 2.0.6 to 2.0.7 in /llama_stack/ui by @dependabot[bot] in #3364
• chore(ui-deps): bump lucide-react from 0.510.0 to 0.542.0 in /llama_stack/ui by @dependabot[bot] in #3363
• chore(ui-deps): bump @radix-ui/react-dropdown-menu from 2.1.14 to 2.1.16 in /llama_stack/ui by @dependabot[bot] in #3361
• chore(github-deps): bump astral-sh/setup-uv from 6.6.0 to 6.6.1 by @dependabot[bot] in #3355
• docs: Update changelog by @terrytangyuan in #3343
• chore(github-deps): bump actions/stale from 9.1.0 to 10.0.0 by @dependabot[bot] in #3352
• chore(github-deps): bump actions/setup-node from 4.4.0 to 5.0.0 by @dependabot[bot] in #3353
• chore(github-deps): bump actions/setup-python from 5.6.0 to 6.0.0 by @dependabot[bot] in #3354
• chore(github-deps): bump actions/checkout from 4.1.7 to 5.0.0 by @dependabot[bot] in #3357
• feat: Add Kubernetes auth provider to use SelfSubjectReview and kubernetes api server by @akram in #2559
• docs: add MongoDB to external provider list by @mohammaddaoudfarooqi in #3369
• feat: Adding OpenAI Prompts API by @franciscojavierarceo in #3319
• fix: environment variable typo in inference recorder error message by @derekhiggins in #3374
• fix: use dataset version 4.0.0 or above by @slekkala1 in #3379
• fix: pre-commit failing by @slekkala1 in #3381
• fix(deps): bump datasets versions for all providers by @ashwinb in #3382

New Contributors

• @are-ces made their first contribution in #3302
• @mohammaddaoudfarooqi made their first contribution in #3369

Full Changelog: v0.2.20...v0.2.21

v0.2.20

Here are some key changes that are coming as part of this release.

Build and Environment

• Environment improvements: fixed env var replacement to preserve types.
• Docker stability: fixed container startup failures for Fireworks AI provider.
• Removed absolute paths in build for better portability.

Features

• UI Enhancements: Implemented file upload and VectorDB creation/configuration directly in UI.
• Vector Store Improvements: Added keyword, vector, and hybrid search inside vector store.
• Added S3 authorization support for file providers.
• SQL Store: Added inequality support to where clause.

Documentation

• Fixed post-training docs.
• Added Contributor Guidelines for creating Internal vs. External providers.

Fixes

• Removed unsupported bfcl scoring function.
• Multiple reliability and configuration fixes for providers and environment handling.

Engineering / Chores

• Cleaner internal development setup with consistent paths.
• Incremental improvements to provider integration and vector store behavior.

What's Changed

• docs: fix post_training docs by @cdoern in #3262
• chore: remove absolute paths by @raghotham in #3263
• docs: Contributor guidelines for creating Internal or External providers by @kelbrown20 in #3111
• feat(UI): Implementing File Upload and VectorDB Creation/Configuration in Playground by @franciscojavierarceo in #3266
• fix(env): env var replacement preserve types by @omertuc in #3270
• fix: docker failing to start container [fireworks-ai] by @slekkala1 in #3267
• chore(dev): add inequality support to sqlstore where clause by @mattf in #3272
• feat(s3 auth): add authorization support for s3 files provider by @mattf in #3265
• feat: implement keyword, vector and hybrid search inside vector stores for PGVector provider by @r3v5 in #3064
• fix: Remove bfcl scoring function as not supported by @slekkala1 in #3281

Full Changelog: v0.2.19...v0.2.20

New Contributors

• @omertuc made their first contribution in #3270
• @r3v5 made their first contribution in vector store hybrid search

v0.2.19

Highlights

• feat: Add CORS configuration support for server by @skamenan7 in #3201
• feat(api): introduce /rerank by @ehhuang in #2940
• feat: Add S3 Files Provider by @mattf in #3202

What's Changed

• chore: Faster npm pre-commit by @franciscojavierarceo in #3206
• fix: disable ui-prettier & ui-eslint by @mattf in #3207
• chore(pre-commit): add pre-commit hook to enforce llama_stack logger usage by @Elbehery in #3061
• fix: fix openai_embeddings for asymmetric embedding NIMs by @jiayin-nvidia in #3205
• chore(files tests): update files integration tests and fix inline::localfs by @mattf in #3195
• fix: Fix broken package-lock.json by @franciscojavierarceo in #3209
• fix: Use pool_pre_ping=True in SQLAlchemy engine creation by @omertuc in #3208
• fix: handle mcp tool calls in previous response correctly by @grs in #3155
• chore: Update dependabot to capture package-lock.json by @franciscojavierarceo in #3212
• chore(python-deps): bump weaviate-client from 4.16.5 to 4.16.9 by @dependabot[bot] in #3219
• chore(python-deps): bump locust from 2.38.0 to 2.39.0 by @dependabot[bot] in #3221
• chore(ui-deps): bump tailwind-merge from 3.3.0 to 3.3.1 in /llama_stack/ui by @dependabot[bot] in #3223
• chore(ui-deps): bump @radix-ui/react-separator from 1.1.6 to 1.1.7 in /llama_stack/ui by @dependabot[bot] in #3222
• chore(ui-deps): bump eslint-config-prettier from 10.1.5 to 10.1.8 in /llama_stack/ui by @dependabot[bot] in #3220
• chore(ui-deps): bump @radix-ui/react-collapsible from 1.1.11 to 1.1.12 in /llama_stack/ui by @dependabot[bot] in #3218
• chore(python-deps): bump chromadb from 1.0.16 to 1.0.20 by @dependabot[bot] in #3217
• chore(ui-deps): bump typescript from 5.8.3 to 5.9.2 in /llama_stack/ui by @dependabot[bot] in #3216
• chore(github-deps): bump actions/setup-node from 4.1.0 to 4.4.0 by @dependabot[bot] in #3214
• chore(github-deps): bump amannn/action-semantic-pull-request from 5.5.3 to 6.1.0 by @dependabot[bot] in #3215
• chore(python-deps): bump llama-api-client from 0.1.2 to 0.2.0 by @dependabot[bot] in #3173
• chore(github-deps): bump actions/checkout from 4.2.2 to 5.0.0 by @dependabot[bot] in #3178
• chore(github-deps): bump astral-sh/setup-uv from 6.4.3 to 6.5.0 by @dependabot[bot] in #3179
• feat: Add CORS configuration support for server by @skamenan7 in #3201
• feat: Remove initialize() Method from LlamaStackAsLibrary by @Elbehery in #2979
• docs: update the docs for NVIDIA Inference provider by @jiayin-nvidia in #3227
• fix: fix the error type in embedding test case by @jiayin-nvidia in #3197
• refactor(logging): rename llama_stack logger categories by @Elbehery in #3065
• feat(UI): Adding Conversation History by @franciscojavierarceo in #3203
• feat(api): introduce /rerank by @ehhuang in #2940
• feat: Add S3 Files Provider by @mattf in #3202
• chore: Add UI linter back by @franciscojavierarceo in #3230
• fix: ensure assistant message is followed by tool call message as expected by openai by @grs in #3224
• chore: indicate to mypy that InferenceProvider.rerank is concrete by @mattf in #3238
• chore: indicate to mypy that InferenceProvider.batch_completion/batch_chat_completion is concrete by @mattf in #3239
• feat: implement query_metrics by @cdoern in #3074
• feat(distro): fork off a starter-gpu distribution by @ashwinb in #3240
• feat: Add optional idempotency support to batches API by @mattf in #3171
• chore(github-deps): bump actions/setup-node from 4.1.0 to 4.4.0 by @dependabot[bot] in #3246
• chore(ui-deps): bump remeda from 2.26.1 to 2.30.0 in /llama_stack/ui by @dependabot[bot] in #3242
• chore(ui-deps): bump @testing-library/dom from 10.4.0 to 10.4.1 in /llama_stack/ui by @dependabot[bot] in #3244
• chore(ui-deps): bump eslint-plugin-prettier from 5.4.0 to 5.5.4 in /llama_stack/ui by @dependabot[bot] in #3241
• chore(github-deps): bump amannn/action-semantic-pull-request from 6.1.0 to 6.1.1 by @dependabot[bot] in #3248
• chore(github-deps): bump astral-sh/setup-uv from 6.5.0 to 6.6.0 by @dependabot[bot] in #3247
• chore(ui-deps): bump @testing-library/jest-dom from 6.6.3 to 6.8.0 in /llama_stack/ui by @dependabot[bot] in #3243
• feat(testing): remove SQLite dependency from inference recorder by @derekhiggins in #3254
• feat: Add example notebook for Langchain + LLAMAStack integration by @slekkala1 in #3228
• chore: Add example notebook for Langchain + LLAMAStack integration (#3228) by @mattf in #3259
• feat(distro): no huggingface provider for starter by @ashwinb in #3258

Full Changelog: v0.2.18...v0.2.19