This release includes 1 security fix for security teams reviewing exposed deployments.
Topics
+4 more
Affected surfaces
Summary
AI summaryBroad release touches New features, Go, fix, and Fix.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Security | Medium |
Fix Langfuse chat observation. Fix Langfuse chat observation. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Security | Medium |
Always use RestrictedUnpickler in deserialize_b64 (CWE-502). Always use RestrictedUnpickler in deserialize_b64 (CWE-502). Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Security | Medium |
Fix SSRF guard for agent test_db_connection endpoint. Fix SSRF guard for agent test_db_connection endpoint. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Adds local & SSH providers in admin panel. Adds local & SSH providers in admin panel. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Feature | Medium |
Agent: structured output aggregation. Agent: structured output aggregation. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Feature | Medium |
Agent: metadata filter reuse. Agent: metadata filter reuse. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Feature | Medium |
Feats: add local & ssh provider in admin panel (duplicate of earlier fact). Feats: add local & ssh provider in admin panel (duplicate of earlier fact). Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Feature | Medium |
Pushes metadata filters down to Infinity, significantly speeding up metadata filtering. Pushes metadata filters down to Infinity, significantly speeding up metadata filtering. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Feats: This enables SelectWithSearch to search by label. Feats: This enables SelectWithSearch to search by label. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Fix session deletion leaking chat-upload blobs. Fix session deletion leaking chat-upload blobs. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Fix: add document download endpoint and refactor existing download function. Fix: add document download endpoint and refactor existing download function. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Go: implement provider: CometAPI. Go: implement provider: CometAPI. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Go: add restful api route aliases. Go: add restful api route aliases. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Go: implement provider: DeepInfra, XunFei. Go: implement provider: DeepInfra, XunFei. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Fix admin CLI system variable commands. Fix admin CLI system variable commands. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Go: implement provider: MinerU. Go: implement provider: MinerU. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Add Replicate chat provider. Add Replicate chat provider. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Feature | Medium |
Go: implement provider: MinerU_Local. Go: implement provider: MinerU_Local. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Performance | Medium |
Accelerated dataset search path, reducing latency by 50–100%. Accelerated dataset search path, reducing latency by 50–100%. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Performance | Medium |
Improved server startup speed and memory usage. Improved server startup speed and memory usage. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Performance | Medium |
Optimizes connector dashboard. Optimizes connector dashboard. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Performance | Medium |
Refactor: enhance graphrag - part 2. Refactor: enhance graphrag - part 2. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Performance | Medium |
Refactor: enhance graphrag - part 2 (duplicate of earlier fact). Refactor: enhance graphrag - part 2 (duplicate of earlier fact). Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Performance | Medium |
Refactor: Drop the vector fetch for ES. Refactor: Drop the vector fetch for ES. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Performance | Medium |
Refactor: speed up ragflow server, save startup memory. Refactor: speed up ragflow server, save startup memory. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Performance | Medium |
Refactor: speed up ragflow server, save startup memory (duplicate of earlier fact). Refactor: speed up ragflow server, save startup memory (duplicate of earlier fact). Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Performance | Medium |
Drops vector fetch for Elasticsearch, speeding up retrieval. Drops vector fetch for Elasticsearch, speeding up retrieval. Source: granite4.1:30b@2026-05-20-audit Confidence: low |
— |
| Bugfix | Medium |
Fix metadata_condition returning all docs when filter matches nothing. Fix metadata_condition returning all docs when filter matches nothing. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: high |
— |
| Bugfix | Medium |
Bugfix: fix tag show. Bugfix: fix tag show. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Bugfix | Medium |
Fix: agent session log message. Fix: agent session log message. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Bugfix | Medium |
Fix discord async issue. Fix discord async issue. Source: granite4.1:8b-q6_K@2026-05-20 Confidence: low |
— |
| Bugfix | Low |
Fixes tag display issue. Fixes tag display issue. Source: granite4.1:30b@2026-05-20-audit Confidence: low |
— |
Full changelog
Summary
Released on May 20, 2026.
New features
- Adds local & SSH providers in admin panel. #15039
Improvements
- Accelerated dataset search path, reducing latency by 50–100% by removing expensive vector fetch and rerank similarity computation steps. #14970
- Pushes metadata filters down to Infinity, significantly speeding up metadata filtering. #14974
- added Redis caching for TTS. #14851
- atomic document counter updates #14867
- Improved server startup speed and memory usage #14973
- Agent: structured output aggregation #13384 #14848
- Agent: metadata filter reuse. #14849
- Optimizes connector dashboard. #14979
What's Changed
- fix: optimize reranking module robustness and bug fixes by @07heco in https://github.com/infiniflow/ragflow/pull/14264
- fix: replace broken assert with raise ValueError in variable_assigner and loop by @Ricardo-M-L in https://github.com/infiniflow/ragflow/pull/13906
- Fix: send input and output token usage to Langfuse by @Br1an67 in https://github.com/infiniflow/ragflow/pull/13294
- Go: implement TTS for MiniMax provider and CLI testing for TTS by @Haruko386 in https://github.com/infiniflow/ragflow/pull/14911
- fix: correct nested path traversal in set_variable_param_value by @Ricardo-M-L in https://github.com/infiniflow/ragflow/pull/13986
- Fix WebDriver resource leak in HTML-to-PDF conversion by @Ricardo-M-L in https://github.com/infiniflow/ragflow/pull/14310
- Doc: code component output section by @buua436 in https://github.com/infiniflow/ragflow/pull/14915
- GO: align time units with Python and centralize timestamp injection in BaseModel by @buua436 in https://github.com/infiniflow/ragflow/pull/14875
- Fix: missing authentication on agent file upload and download endpoints by @dale053 in https://github.com/infiniflow/ragflow/pull/14854
- Go: fix LastLoginTime update by @buua436 in https://github.com/infiniflow/ragflow/pull/14917
- fix: replace mutable default arguments with None in LLM chat models by @gambletan in https://github.com/infiniflow/ragflow/pull/13513
- Go: update user settings fields by @buua436 in https://github.com/infiniflow/ragflow/pull/14918
- fix: atomic chunk/token counter updates for documents and knowledge b… by @dale053 in https://github.com/infiniflow/ragflow/pull/14867
- fix(agent): handle duplicate MCP tool names by @eviaaaaa in https://github.com/infiniflow/ragflow/pull/14217
- fix: close db connections reliably in test_db_connection by @wdeveloper16 in https://github.com/infiniflow/ragflow/pull/14777
- Go: implement TTS for fishaudio, openrouter and asr for fishaudio by @Haruko386 in https://github.com/infiniflow/ragflow/pull/14926
- Feat: This enables SelectWithSearch to search by label. by @cike8899 in https://github.com/infiniflow/ragflow/pull/14925
- Doc: Finalized v0.25.4 release notes by @writinwaters in https://github.com/infiniflow/ragflow/pull/14929
- fix: add document download endpoint and refactor existing download function by @buua436 in https://github.com/infiniflow/ragflow/pull/14927
- fix: preserve uploaded file attachments after subsequent assistant messages by @octo-patch in https://github.com/infiniflow/ragflow/pull/13993
- fix(llm): Tongyi-Qianwen embeddings use correct DashScope native API for intl URLs by @sham-sr in https://github.com/infiniflow/ragflow/pull/14784
- fix(agent): pass top_k and fix similarity weight slider behavior by @yingjianzh in https://github.com/infiniflow/ragflow/pull/14760
- security: always use RestrictedUnpickler in deserialize_b64 (CWE-502) by @sebastiondev in https://github.com/infiniflow/ragflow/pull/14803
- Go: add file parse command by @JinHai-CN in https://github.com/infiniflow/ragflow/pull/14892
- fix(agentbot): aggregate structured output in non-streaming completions by @plind-junior in https://github.com/infiniflow/ragflow/pull/14848
- fix: enable GitHub connector to sync PRs and issues by default by @octo-patch in https://github.com/infiniflow/ragflow/pull/14062
- Go: implement Rerank in vLLM driver (#14878) by @hunnyboy1217 in https://github.com/infiniflow/ragflow/pull/14880
- docs: add FAQ entry for using Ollama with RAGFlow by @SnakeEye-sudo in https://github.com/infiniflow/ragflow/pull/14557
- Go: add cli command, list dataset documents by @JinHai-CN in https://github.com/infiniflow/ragflow/pull/14948
- Go: implement TTS, ASR for Siliconflow and TTs for StepFun by @Haruko386 in https://github.com/infiniflow/ragflow/pull/14944
- fix: correct attribute name typo model_speciess to model_species by @Ricardo-M-L in https://github.com/infiniflow/ragflow/pull/13929
- feat: bump Python minimum from 3.12 to 3.13, drop strenum backport by @wdeveloper16 in https://github.com/infiniflow/ragflow/pull/14767
- Go: implement PaddleOCR provider and implement ASR for CoHere by @Haruko386 in https://github.com/infiniflow/ragflow/pull/14954
- Feat/web markdown UI updates by @xjq123456 in https://github.com/infiniflow/ragflow/pull/14214
- fix: correct literal_eval dispatch and bool isinstance ordering in agent components by @Ricardo-M-L in https://github.com/infiniflow/ragflow/pull/13988
- refactor: optimize BaseTitleChunker to improve RAG document chunk quality by @07heco in https://github.com/infiniflow/ragflow/pull/14247
- Bump to infinity v0.7.0 by @qinling0210 in https://github.com/infiniflow/ragflow/pull/14968
- Fix session deletion leaking chat-upload blobs by @6ba3i in https://github.com/infiniflow/ragflow/pull/14969
- Go: implement Embed (embeddings) in Novita driver by @pandadev66 in https://github.com/infiniflow/ragflow/pull/14895
- Go: add Jina chat completions support by @jakearmstrong59 in https://github.com/infiniflow/ragflow/pull/14935
- Add Anthropic Go model provider by @carlos4s in https://github.com/infiniflow/ragflow/pull/14940
- Fix: guard empty/whitespace embedding inputs in LLMBundle (#14428) by @Rene0422 in https://github.com/infiniflow/ragflow/pull/14924
- Refactor: Drop the vector fetch for ES by @KevinHuSh in https://github.com/infiniflow/ragflow/pull/14970
- Push metadata filters down to Infinity by @qinling0210 in https://github.com/infiniflow/ragflow/pull/14974
- Go: implement provider: CometAPI by @tmimmanuel in https://github.com/infiniflow/ragflow/pull/14930
- Fix: add SSRF guard for agent test_db_connection endpoint by @dale053 in https://github.com/infiniflow/ragflow/pull/14860
- Go: fix siliconflow embedding response by @buua436 in https://github.com/infiniflow/ragflow/pull/14975
- Refactor: speed up ragflow server, save startup memory by @wangq8 in https://github.com/infiniflow/ragflow/pull/14973
- fix(api): enforce tenant access for connector routes by @dev111-actor in https://github.com/infiniflow/ragflow/pull/14747
- Refactor: enhance graphrag - part 2 by @wangq8 in https://github.com/infiniflow/ragflow/pull/14972
- Go: add restful api route aliases by @buua436 in https://github.com/infiniflow/ragflow/pull/14977
- Go: implement provider: DeepInfra, XunFei by @Haruko386 in https://github.com/infiniflow/ragflow/pull/14978
- fix: metadata_condition returning all docs when filter matches nothing by @hak2979 in https://github.com/infiniflow/ragflow/pull/14967
- Bugfix: fix tag show by @wangq8 in https://github.com/infiniflow/ragflow/pull/14980
- Fix admin CLI system variable commands by @jakearmstrong59 in https://github.com/infiniflow/ragflow/pull/14956
- Go: fix forgetting policy validation and fix memory update diff checks by @buua436 in https://github.com/infiniflow/ragflow/pull/14976
- Go: align document list response by @buua436 in https://github.com/infiniflow/ragflow/pull/14982
- Go: add dataset graph api by @buua436 in https://github.com/infiniflow/ragflow/pull/14984
- Feat: full optimization on connector dashboard by @Magicbook1108 in https://github.com/infiniflow/ragflow/pull/14979
- Fix: validate memory tenant model IDs on update and enforce tenant scope in memory pipeline by @jony376 in https://github.com/infiniflow/ragflow/pull/14923
- Misc: fix some typos by @kingloon in https://github.com/infiniflow/ragflow/pull/14987
- Go: implement provider: MinerU by @Haruko386 in https://github.com/infiniflow/ragflow/pull/14990
- Add Replicate chat provider by @tmimmanuel in https://github.com/infiniflow/ragflow/pull/14958
- Fix: agent session log message by @buua436 in https://github.com/infiniflow/ragflow/pull/14991
- Fix: MinerU vlm-http-client backend output file detection by @lksr1201 in https://github.com/infiniflow/ragflow/pull/14240
- fix: the opencc c library uses fgets() to read dicti... in text.c by @orbisai0security in https://github.com/infiniflow/ragflow/pull/13970
- feat(tts): cache synthesized speech in Redis to avoid redundant calls by @plind-junior in https://github.com/infiniflow/ragflow/pull/14851
- fix(retrieval): keep manual metadata filter reusable inside Iteration by @plind-junior in https://github.com/infiniflow/ragflow/pull/14849
- Go: implement provider: Xinference by @tmimmanuel in https://github.com/infiniflow/ragflow/pull/14938
- Add TogetherAI chat provider by @tmimmanuel in https://github.com/infiniflow/ragflow/pull/14957
- Fix: restore saved api_key fallback in add_llm (#14921) by @Rene0422 in https://github.com/infiniflow/ragflow/pull/14941
- Feat: add new tests and tescases for restful api suite by @6ba3i in https://github.com/infiniflow/ragflow/pull/14993
- Feat: VLM image descriptions in MinerU parser (#14869) by @Rene0422 in https://github.com/infiniflow/ragflow/pull/14946
- feat(sdk): make Begin inputs discoverable on Session.ask by @plind-junior in https://github.com/infiniflow/ragflow/pull/14842
- Feat: add new tests and tescases for restful api suite by @6ba3i in https://github.com/infiniflow/ragflow/pull/14996
- Refact functions in engine in GO by @qinling0210 in https://github.com/infiniflow/ragflow/pull/14981
- Go: implement provider: 302.AI and JieKou-AI by @Haruko386 in https://github.com/infiniflow/ragflow/pull/15034
- Feat: add new tests and tescases for restful api suite by @6ba3i in https://github.com/infiniflow/ragflow/pull/15038
- fix(go): guard custom base URL driver creation by @bitloi in https://github.com/infiniflow/ragflow/pull/15030
- Fix: The folder tree menu for moving folders cannot be scrolled. by @cike8899 in https://github.com/infiniflow/ragflow/pull/15037
- fix: Langfuse chat observation by @bitloi in https://github.com/infiniflow/ragflow/pull/15026
- Feat: add local & ssh provider in admin panel by @Magicbook1108 in https://github.com/infiniflow/ragflow/pull/15039
- Fix discord async issue by @wangq8 in https://github.com/infiniflow/ragflow/pull/15054
- Go: implement provider: MinerU_Local by @Haruko386 in https://github.com/infiniflow/ragflow/pull/15051
- Docs: Initial draft of v0.25.5 release notes. by @writinwaters in https://github.com/infiniflow/ragflow/pull/15058
- Docs: Update version references to v0.25.5 in READMEs and docs by @JinHai-CN in https://github.com/infiniflow/ragflow/pull/15059
New Contributors
- @Br1an67 made their first contribution in https://github.com/infiniflow/ragflow/pull/13294
- @sham-sr made their first contribution in https://github.com/infiniflow/ragflow/pull/14784
- @yingjianzh made their first contribution in https://github.com/infiniflow/ragflow/pull/14760
- @SnakeEye-sudo made their first contribution in https://github.com/infiniflow/ragflow/pull/14557
- @xjq123456 made their first contribution in https://github.com/infiniflow/ragflow/pull/14214
- @jakearmstrong59 made their first contribution in https://github.com/infiniflow/ragflow/pull/14935
- @carlos4s made their first contribution in https://github.com/infiniflow/ragflow/pull/14940
- @Rene0422 made their first contribution in https://github.com/infiniflow/ragflow/pull/14924
- @dev111-actor made their first contribution in https://github.com/infiniflow/ragflow/pull/14747
- @hak2979 made their first contribution in https://github.com/infiniflow/ragflow/pull/14967
- @kingloon made their first contribution in https://github.com/infiniflow/ragflow/pull/14987
- @lksr1201 made their first contribution in https://github.com/infiniflow/ragflow/pull/14240
Full Changelog: https://github.com/infiniflow/ragflow/compare/v0.25.4...v0.25.5
Breaking Changes
- Bumped minimum required Python version from 3.12 to 3.13
Security Fixes
- Always use RestrictedUnpickler in deserialize_b64 to mitigate CWE-502 deserialization vulnerability
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About ragflow
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
Beta — feedback welcome: [email protected]