Open source
Libraries, tools and repos: releases you depend on and projects gaining stars.
llm-openai-decisions 0.1a0
OpenAI released their new Jev-style Decisions API, as previously announced at last week's DevDay.
langchain-ai/langchain langchain-openai==1.7.0
chore(deps): bump langgraph-sdk from 0.4.4 to 0.4.6 in /libs/partners/openai (#41096)
anthropics/anthropic-sdk-python v1.13.0
api: add types for the Chat and Cowork unified analytics metrics
crewAIInc/crewAI 1.15.27
Add deepinfra as an OpenAI-compatible provider
BerriAI/litellm v1.105.0
Verify using the pinned commit hash (recommended):
huggingface/diffusers v0.41.0: Diffusers 0.41.0: QwenImage 2.1 pipeline and more
> This release brings Qwen-Image 2.1 to Diffusers, with text-to-image generation, image editing, native transparency, and LoRA training.
LegalOn halves Codex costs while maintaining development speed
LegalOn cut estimated daily Codex costs by 65% while maintaining development speed.
ollama/ollama v0.40.2
Models downloaded with earlier versions of Ollama are upgraded in the background the first time you run them, for better performance and compatibility when running on llama.cpp.
pydantic/pydantic-ai v2.54.0: v2.54.0 (2026-10-02)
<!-- Release notes generated using configuration in .github/release.yml at main -->
[AINews] Pi 1.0, Pi Durable, and AIE NYC
Last call for regular tickets for AI Engineer NYC!
langchain-ai/langgraph sdk==0.4.6: langgraph-sdk==0.4.6
fix(sdk-py): percent-encode threadid and assistantid in thread stream requests (#9213)
AI Native by Design: Lessons Learned from Building NVIDIA TensorRT Model Connect
Parallel work, model-family isolation, reversible changes, and GPU-backed validation shaped an open source project designed around coding agents NVIDIA TensorRT...Parallel work, model-family isolation, reversible changes, and GPU-backed validation shaped an open source project designed around coding agents NVIDIA TensorRT Model Connect is an open source collection of AI model reference implementations in C++, built on top of NVIDIA TensorRT.
NVIDIA/TensorRT-LLM v1.3.0rc29
Expose Nemotron-H vision-language LoRA configuration for supported inference paths #19151
modelcontextprotocol/python-sdk v2.3.0
A few things behave differently, so skim these first:
ray-project/ray ray-2.59.0: Ray-2.59.0
Ray Data LLM & Ray Serve LLM are GA/Stable: the LLM APIs graduate to general availability this release (\#65194), alongside an upgrade to vLLM 0.27.0 (\#65351).
kkjiangk/agent-py: Personal AIOps engineering project with durable workflows, scoped hybrid retrieval, reliable event delivery, and verified API/browser contra
Personal AIOps engineering project with durable workflows, scoped hybrid retrieval, reliable event delivery, and verified API/browser contracts.
vllm-project/vllm v0.30.0
This release features 762 commits from 315 contributors (104 new)!
modelcontextprotocol/typescript-sdk v2.3.1: 2.3.1
requireBearerAuth in @modelcontextprotocol/server-legacy takes the optional expectedResource that @modelcontextprotocol/server 2.3.0 added: it accepts only tokens issued for this server (the token's audience).
Together Link: open models in the harness you already use. Start with one command today.
Together Link brings frontier open models like GLM 5.3 and Kimi K3 into the coding agent your team already uses, cutting model spend by over 50%.
kōdlo
End-to-end sales workspace with unified context
Python 3.15.0 added to actions/python-versions
Now that this has landed, you can add "3.15" to a GitHub Actions testing matrix to run tests against the new Python 3.15.0 release.
anthropics/anthropic-sdk-python v1.12.1
docs: note that listing Claude Console spend limits is in early access
langchain-ai/langchain langchain==1.4.4
chore(deps): bump langgraph-sdk from 0.4.2 to 0.4.4 in /libs/langchainv1 (#41074)
langchain-ai/langchain langchain-core==1.6.9
feat(core): accept a callable in withretry (#41158)
langchain-ai/langchain langchain-core==1.6.8
fix(core): harden scoped and compatible IPv6 SSRF checks (#41151)
BerriAI/litellm v1.104.3
Verify using the pinned commit hash (recommended):
ttok 0.4
ttok is my CLI tool for counting tokens, using OpenAI's open source tiktoken library.
langchain-ai/langchain langchain-fireworks==1.7.1
chore(deps): bump langgraph-sdk from 0.4.4 to 0.4.6 in /libs/partners/fireworks (#41100)
Can a Cloud-Native Harness Make Agents Reliable Beyond the Desktop?
Since that’s easiest to implement locally, many coding agents started out as terminal tools and then desktop apps (Claude Code famously began as a CLI tool).
langchain-ai/langchain langchain-core==1.6.7
fix(core): include openai redactedcontent in v1 output for bedrock converse (#41086)
crewAIInc/crewAI 1.15.26
Fix ownership retention of rwlock when acquisition is interrupted.
pydantic/pydantic-ai v2.52.0: v2.52.0 (2026-09-29)
This release fixes one security issue in webfetch.
BerriAI/litellm v1.104.2
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.102.4
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.101.6
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.104.1
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.103.4
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.102.3
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.100.5
Verify using the pinned commit hash (recommended):
BerriAI/litellm v1.101.5
Verify using the pinned commit hash (recommended):
crewAIInc/crewAI 1.15.24
Add crewai eval to print a markdown brief when run by an agent
BerriAI/litellm v1.104.0
Verify using the pinned commit hash (recommended):
langchain-ai/langchain langchain-openai==1.6.7
chore(model-profiles): refresh model profile data (#40924)
Deno is joining Cloudflare
The Deno team released the first version of celld back in August - their open source implementation of the Durable Objects pattern from Cloudflare Workers.
ollama/ollama v0.35.0
Decision models return choices, probabilities, and scores instead of text.
langchain-ai/langchain langchain-anthropic==1.7.5
fix(anthropic): support Claude Sonnet 5.5 compatibility (#40882)
langchain-ai/langchain langchain-text-splitters==1.1.3
chore(deps): bump tornado from 6.5.9 to 6.5.10 in /libs/text-splitters (#40971)
langchain-ai/langchain langchain-fireworks==1.7.0
hotfix(fireworks): replace unavailable integration test model (#40890)