Dify: v1.13.2
Dify 1.13.2. Source highlight: This patch release fixes several critical regressions and stability issues introduced in v1.13.1, including:
AI Models & LLM Frameworks archive — Page 6
1172 tracked entries. Older history is imported progressively; see source coverage.
Ollama v0.18.2 — Release Notes & Engineering Impact
Ollama v0.18.2. Source highlight: Add extra check to ensure `npm` and `git` are installed before installing OpenClaw
Dify: v1.13.1
Dify 1.13.1. Source highlight: 🚀 Major Functional Changes
Ollama v0.18.1 — Release Notes & Engineering Impact
Ollama v0.18.1. Source highlight: Web Search and Fetch in OpenClaw
LocalAI v4.0.0 — Release Notes & Engineering Impact
LocalAI v4.0.0. Source highlight: 🎉 LocalAI 4.0.0 Release! 🚀
Ollama v0.18.0 — Release Notes & Engineering Impact
Ollama v0.18.0. Source highlight: Ollama 0.18 includes improved performance for OpenClaw and Ollama’s cloud models, including the new Nemotron-3-Super model by NVIDIA designed for high-performance agentic reasoning tasks.
LangChain langchain==1.2.12 — Release Notes & Engineering Impact
LangChain langchain==1.2.12. Source highlight: Changes since langchain==1.2.11
vLLM v0.17.1 — Release Notes & Engineering Impact
vLLM v0.17.1. Source highlight: This is a patch release on top of `v0.17.0` to address a few issues:
LangChain langchain==1.2.11 — Release Notes & Engineering Impact
LangChain langchain==1.2.11. Source highlight: Changes since langchain==1.2.10
Open WebUI v0.8.10 — Release Notes & Engineering Impact
Open WebUI v0.8.10. Source highlight: 🔐 **Custom OIDC logout endpoint.** Administrators can now configure a custom OpenID Connect logout URL via OPENID_END_SESSION_ENDPOINT, enabling logout functionality for OIDC providers that require custom endpoints like AWS Cognito. Commit
Open WebUI v0.8.9 — Release Notes & Engineering Impact
Open WebUI v0.8.9. Source highlight: ▶️ **Open Terminal notebook cell execution.** Users can now run Jupyter Notebook code cells directly in the Open Terminal file navigator, execute entire notebooks with a single click, edit and modify cells before running, and control the kernel - bringing full interactive notebook execution to the b
vLLM v0.17.0 — Release Notes & Engineering Impact
vLLM v0.17.0. Source highlight: **Known Issue**: If you are on CUDA 12.9+ and encounter a `CUBLAS_STATUS_INVALID_VALUE` error, this is caused by a CUDA library mismatch. To resolve, try one of the following:
LangChain langchain==0.3.28 — Release Notes & Engineering Impact
LangChain langchain==0.3.28. Source highlight: Changes since langchain==0.3.27
Ollama v0.17.7 — Release Notes & Engineering Impact
Ollama v0.17.7. Source highlight: Allow thinking levels such as `"medium"` to correctly interpreted in Ollama's API for all thinking models
Hugging Face Transformers: v5.3.0: EuroBERT, VibeVoice ASR, TimesFM2.5, PP-DocLayoutV2, OlmoHybrid, ModernVBert, Higgs Audio V2
Hugging Face Transformers v5.3.0. Source highlight: EuroBERT is a multilingual encoder model based on a refreshed transformer architecture, akin to Llama but with bidirectional attention. It supports a mixture of European and widely spoken languages, with sequences of up to 8192 tokens.
Ollama v0.17.6 — Release Notes & Engineering Impact
Ollama v0.17.6. Source highlight: Fixed issue where GLM-OCR would not work due to incorrect prompt rendering
Open WebUI v0.8.8 — Release Notes & Engineering Impact
Open WebUI v0.8.8. Source highlight: [0.8.8] - 2026-03-02
Ollama v0.17.5 — Release Notes & Engineering Impact
Ollama v0.17.5. Source highlight: Qwen3.5: the small Qwen 3.5 model series is now available in 0.8B, 2B, 4B and 9B parameter sizes.
Open WebUI v0.8.7 — Release Notes & Engineering Impact
Open WebUI v0.8.7. Source highlight: [0.8.7] - 2026-03-01
Open WebUI v0.8.6 — Release Notes & Engineering Impact
Open WebUI v0.8.6. Source highlight: [0.8.6] - 2026-03-01
Ollama v0.17.4 — Release Notes & Engineering Impact
Ollama v0.17.4. Source highlight: Qwen 3.5: a family of open-source multimodal models that delivers exceptional utility and performance.
Ollama v0.17.3 — Release Notes & Engineering Impact
Ollama v0.17.3. Source highlight: Fixed issue where tool calls in the Qwen 3 and Qwen 3.5 model families would not be parsed correctly if emitted during thinking
Ollama v0.17.2 — Release Notes & Engineering Impact
Ollama v0.17.2. Source highlight: Fixed issue where Ollama's app on Windows would crash when a new update has been downloaded
vLLM v0.16.0 — Release Notes & Engineering Impact
vLLM v0.16.0. Source highlight: Please note that this release was branch cut on Feb 8, so any features added to vLLM after that date is not included.
Ollama v0.17.1 — Release Notes & Engineering Impact
Ollama v0.17.1. Source highlight: Nemotron architecture support in Ollama's engine
Open WebUI v0.8.5 — Release Notes & Engineering Impact
Open WebUI v0.8.5. Source highlight: [0.8.5] - 2026-02-23
Open WebUI v0.8.4 — Release Notes & Engineering Impact
Open WebUI v0.8.4. Source highlight: [0.8.4] - 2026-02-23
LocalAI v3.12.1 — Release Notes & Engineering Impact
LocalAI v3.12.1. Source highlight: This is a patch release to tag the new llama.cpp version which fixes incompatibilities with Qwen 3 coder.
Ollama v0.17.0 — Release Notes & Engineering Impact
Ollama v0.17.0. Source highlight: OpenClaw can now be installed and configured automatically via Ollama, making it the easiest way to get up and running with OpenClaw with open models like Kimi-K2.5, GLM-5, and Minimax-M2.5.
LocalAI v3.12.0 — Release Notes & Engineering Impact
LocalAI v3.12.0. Source highlight: 🎉 LocalAI 3.12.0 Release! 🚀
Ollama v0.16.3 — Release Notes & Engineering Impact
Ollama v0.16.3. Source highlight: New `ollama launch cline` added for the Cline CLI
Open WebUI v0.8.3 — Release Notes & Engineering Impact
Open WebUI v0.8.3. Source highlight: [0.8.3] - 2026-02-17
Hugging Face Transformers: v5.2.0: GLM-5, Qwen3.5, Voxtral Realtime, VibeVoice Acoustic Tokenizer
Hugging Face Transformers v5.2.0. Source highlight: VoxtralRealtime is a streaming speech-to-text model from Mistral AI, designed for real-time automatic speech recognition (ASR). Unlike the offline Voxtral model which processes complete audio files, VoxtralRealtime is architected for low-latency, incremental transcription by processing audio in chun
Open WebUI v0.8.2 — Release Notes & Engineering Impact
Open WebUI v0.8.2. Source highlight: [0.8.2] - 2026-02-16
Ollama v0.16.2 — Release Notes & Engineering Impact
Ollama v0.16.2. Source highlight: `ollama launch claude` now supports searching the web when using `:cloud` models
Open WebUI v0.8.1 — Release Notes & Engineering Impact
Open WebUI v0.8.1. Source highlight: [0.8.1] - 2026-02-14
Open WebUI v0.8.0 — Release Notes & Engineering Impact
Open WebUI v0.8.0. Source highlight: [0.8.0] - 2026-02-12
Ollama v0.16.1 — Release Notes & Engineering Impact
Ollama v0.16.1. Source highlight: Installing Ollama via the `curl` install script on macOS will now only prompt for your password if its required
Ollama v0.16.0 — Release Notes & Engineering Impact
Ollama v0.16.0. Source highlight: GLM-5: A strong reasoning and agentic model from Z.ai with 744B total parameters (40B active), built for complex systems engineering and long-horizon tasks.
Dify: 1.13.0 - Human-in-the-Loop and Workflow Execution Upgrades
Dify 1.13.0. Source highlight: Human-in-the-Loop (HITL)
LangChain langchain==1.2.10 — Release Notes & Engineering Impact
LangChain langchain==1.2.10. Source highlight: Changes since langchain==1.2.9
LocalAI v3.11.0 — Release Notes & Engineering Impact
LocalAI v3.11.0. Source highlight: 🎉 LocalAI 3.11.0 Release! 🚀
Ollama v0.15.6 — Release Notes & Engineering Impact
Ollama v0.15.6. Source highlight: Fixed context limits when running `ollama launch droid`
LangChain langchain==1.2.9 — Release Notes & Engineering Impact
LangChain langchain==1.2.9. Source highlight: Changes since langchain==1.2.8
Hugging Face Transformers: v5.1.0: EXAONE-MoE, PP-DocLayoutV3, Youtu-LLM, GLM-OCR
Hugging Face Transformers v5.1.0. Source highlight: K-EXAONE is a large-scale multilingual language model developed by LG AI Research. Built using a Mixture-of-Experts architecture, K-EXAONE features 236 billion total parameters, with 23 billion active during inference. Performance evaluations across various benchmarks demonstrate that K-EXAONE excel
vLLM v0.15.1 — Release Notes & Engineering Impact
vLLM v0.15.1. Source highlight: v0.15.1 is a patch release with security fixes, RTX Blackwell GPU fixes support, and bug fixes.
Dify: v1.12.1 - Fix base URL in client and other bugs
Dify 1.12.1. Source highlight: fix(docker): persist IRIS data across container recreation using Durable %SYS
Dify: v1.12.0 - Introducing Summary Index: Smarter Retrieval with AI Summarization
Dify 1.12.0. Source highlight: 🚀 New Features: Summary Index
Ollama v0.15.5 — Release Notes & Engineering Impact
Ollama v0.15.5. Source highlight: Qwen3-Coder-Next: a coding-focused language model from Alibaba's Qwen team, optimized for agentic coding workflows and local development.
LangChain langchain==1.2.8 — Release Notes & Engineering Impact
LangChain langchain==1.2.8. Source highlight: Changes since langchain==1.2.7