// Release stream · daily archive

Wed, Sep 23

108 releases from tracked repos.

Next day: 09-24 → ← Previous: 09-22

Show pre-releases & patches

Wed, Sep 23· 78

nexu-io/open-design open-design-v0.24.0 Minor Desktop app

🌟 Codename: *Every Run, Kept Intact*

00:51
debpalash/VoiceStudio v0.5.6 Patch Desktop app

VoiceStudio now talks to your other tools.** Answer Twilio phone calls in a saved voice, and connect Claude Code, Cursor, Codex CLI and the OpenAI Agents SDK to VoiceStudio with copyable setup that works, including in Do

16:53
unslothai/unsloth v0.1.815-beta Patch Desktop app +2 more

You can now run Qwen-Image-2.1 locally with Unsloth! This release also includes custom Agent Skills, and easier chat/project management. It also brings 2x faster reasoning blocks (60 FPS vs 30 FPS), more reliable trainin

22:51
zed-industries/zed v1.21.0 Minor New model

This week's release includes faster rendering in syntax-highlighted files and Markdown code blocks, a language server command picker, a setting that keeps your system awake during long-running agent turns, and BYOK suppo

23:42
OpenHands/OpenHands v1.22.0 Minor

(2026-09-22)

00:28
langchain-ai/deepagents deepagents==0.7.18 First release

Large tool result previews now more clearly explain when output was clipped or truncated, including byte-cap losses, per-line clipping, and showing truncation notices only when truncation occurs. Legacy large-result prev

11:06
anthropics/claude-code v2.1.280 Patch New agent

Added Claude Opus 5.5 (`claude-opus-5-5`), now the default Opus model — 1M context, $4/$20 per Mtok with $0.20/Mtok cache reads

00:38
twentyhq/twenty sdk/v2.41.0 Sub-package

⚠️ Breaking changes

21:36
aaif-goose/goose v1.52.0 Minor

an open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM

23:10
ChromeDevTools/chrome-devtools-mcp chrome-devtools-mcp-v1.10.0 Minor +1 more

1.10.0 (2026-09-23)

17:52
Gentleman-Programming/gentle-ai v3.6.0 Minor New agent +2 more

Gentle AI v3.6.0 is a minor release with one new capability and three fixes. `gentle-ai install` and `sync` now honor `PI_CODING_AGENT_DIR`, the same override Pi reads, so a Gentle Shell isolated home can be provisioned

01:26
openai/openai-python v3.19.0 Minor +1 more

3.19.0 (2026-09-22)

08:08
mastra-ai/mastra @mastra/[email protected] Sub-package

MCP v2 (2026-07-28) server architecture in `@mastra/[email protected]`

16:40
OtterMind/Chat2DB v5.4.0-beta.1 Minor

Chat2DB Community 5.4.0-beta.1 — early preview

13:14
modelcontextprotocol/typescript-sdk @modelcontextprotocol/[email protected] Sub-package +4 more

The official TypeScript SDK for Model Context Protocol servers and clients

23:42
codewhale-hq/Codewhale v0.10.0 Minor

Open-source Rust agent engine and terminal client for Codewhale, with provider choice, tools, approvals and receipts.

01:28
CopilotKit/CopilotKit channels/v0.11.0 Sub-package +2 more

fix(channels): widen the express floor so an older express stops failing installs (#7331) (1504af8)

01:18
TencentCloud/Octop v1.0.2b1 First release +1 more

A smarter, self-hosted AI assistant — multi-user, multi-agent.

11:00
pydantic/pydantic-ai v2.48.0 Minor

Support `genai-prices` 0.1.7 and new OpenAI audio models by @adtyavrdhn in https://github.com/pydantic/pydantic-ai/pull/8617

10:41
google-deepmind/mujoco 3.14.0 Minor

Multi-Joint dynamics with Contact. A general purpose physics simulator.

00:52
casdoor/casdoor v4.8.0 Minor

eff25f63b7e6231b0d57d43330a5e8a5582e4266 fix: allow an org-scoped subject to update its own organization

14:19
microsoft/onnxruntime plugin-ep-webgpu/v0.4.0 Sub-package

This release brings kernel performance improvements, expanded operator support, and reliability fixes to the ONNX Runtime WebGPU Plugin EP.

01:13
NVIDIA/TensorRT v11.3 Minor

NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.

03:39
NVIDIA-NeMo/Switchyard v0.3.0 Minor New agent

Switchyard v0.3.0

02:06
jo-inc/camofox-browser v1.17.0 Minor

Stealth headless browser for AI agents — bypass Cloudflare, bot detection, and anti-scraping. Drop-in Puppeteer/Playwright replacement.

07:39
Kilo-Org/kilocode v7.7.9 Patch New agent +2 more

Kilo is the all-in-one agentic engineering platform. Build, ship, and iterate faster with the most popular open source coding agent.

19:00
langchain-ai/openwiki v0.6.0 Minor

OpenWiki is a CLI that writes and maintains agent documentation for your codebase.

22:36
kyutai-labs/pocket-tts v3.2.0 Minor

The french language has been updated, we now have `"french"` and a better `"french_24l"`!

23:47
ggml-org/ggml v0.25.0 Minor

Tensor library for machine learning

16:33
langflow-ai/langflow v1.12.3 Patch

✨ New Features

05:22
langchain-ai/langchain langchain-openai==1.6.5 Release

Changes since langchain-openai==1.6.4

23:31
palmier-io/palmier-pro v0.10.0 Minor

https://palmier.io/changelog/v0.10.0

10:20
zhukunpenglinyutong/desktop-cc-gui v1.0.7 First release +1 more

Multi-engine AI coding desktop client (Tauri). Claude Code, Codex, Gemini, OpenCode, DeepSeek Harness and more in one GUI.

01:06
MoonshotAI/kimi-code @moonshot-ai/[email protected] Sub-package

Kimi Code CLI — The Starting Point for Next-Gen Agents

20:31
simonw/llm 0.36 Minor

New features:

02:48
skypilot-org/skypilot v0.14.0rc1 Minor

SkyPilot v0.14.0rc1

07:35
stablyai/orca v1.4.209 Patch

Thank you so much for using Orca and for your continued support! ❤️

13:25
unclecode/crawl4ai v0.9.4 Patch +1 more

Open-source web crawler and scraper for LLMs and AI agents: any website into clean, LLM-ready Markdown. Run it yourself, or use Crawl4AI Cloud with one key. Join our Discord htt…

20:14
Mintplex-Labs/anything-llm v1.16.2 Patch

Stop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience

03:06
anthropics/anthropic-sdk-python v1.8.0 Minor

(2026-09-22)

00:25
bojieli/ai-agent-book latest Rolling

《深入理解 AI Agent：设计原理与工程实践》（李博杰著）开源主仓库：全书正文、编译版 PDF 与按章配套代码

20:58
multica-ai/multica v0.5.2 Patch

af62622e50e5d0cb5d3dfc8d0701311e14c21dca MUL-7288: fix(server): wake guest squad leaders on worker replies (#8314)

18:15
aipoch/open-science v0.33.0 Minor

AIPOCH Open-Science v0.33.0

13:59
Imbad0202/academic-research-skills v3.22.1 Patch

Model currency and repairs; the new prompt-level guard is unmeasured:** v3.22.1 names Claude Opus 5.5 beside Claude Fable 5.1 as a supported session model, after a two-reader audit of the Opus 5.5 system card that retire

13:53
Agenta-AI/agenta v0.120.0 Minor

[feat] Run agent HTML apps from the drive by @ardaerzin in https://github.com/Agenta-AI/agenta/pull/6972

16:55
agno-agi/agno v3.0.11 Patch

New Features:

21:14
TencentCloud/CubeSandbox v0.7.2-rc2 First releasepre

Instant, Concurrent, Secure & Lightweight Sandbox for AI Agents.

12:54
vllm-project/llm-compressor 0.14.0 Minor

State-of-the-art LLM compression, built for production inference with vLLM

02:59
huggingface/pytorch-image-models v1.0.30 Patch New model

September 22, 2026

04:45
superlinked/sie v0.8.0 Minor

Open-source inference server and production cluster for all the models your agent needs.

22:28
Infisical/infisical v0.165.16 Patch

feat: block certificate deletion until expiry by @carlosmonastyrski in https://github.com/Infisical/infisical/pull/8213

23:44
LibreChat-AI/LibreChat v0.8.8-rc4 Patchpre New agent

https://www.librechat.ai/changelog/v0.8.8-rc4

11:58
mlflow/mlflow v2.11.5 Patch

The open source AI engineering platform for agents, LLMs, and ML models. MLflow enables teams of all sizes to debug, evaluate, monitor, and optimize production-quality AI applic…

14:17
Skyvern-AI/skyvern v1.0.54 Patch

Version bumped from 1.0.53 to 1.0.54

06:57
NVIDIA/open-gpu-kernel-modules 595.104.02 Patch

NVIDIA Linux open GPU kernel module source

01:02
coder/coder v2.37.3 Patch

This is a mainline Coder release. We advise enterprise customers without a staging environment to install our latest stable release while we refine this version. Learn more about our Release Schedule.

04:50
triggerdotdev/trigger.dev v4.6.4 Patch

trigger.dev v4.6.4

01:27
XiaomiMiMo/MiMo-Code v0.1.14 Patch +1 more

MiMo Code: Where Models and Agents Co-Evolve

16:11
cloudwego/eino v0.9.21 Patch

feat: expose cache write tokens in model usage by @mrh997 in https://github.com/cloudwego/eino/pull/1309

16:57
run-llama/liteparse node-v2.14.7 Patch +3 more

A fast, helpful, and open-source document parser

05:41
MemTensor/MemOS v2.0.34 Patch

MemoryOS 2.0.34 release.

13:24
Tencent/BrowserSkill dsh-plugin-v0.3.1 Patch +2 more

Let AI agents use your real, logged-in browser without interrupting your work. CLI + extension for browser automation across any shell-capable AI agent.

16:51
mark3labs/mcp-go v1.1.1 Patch

A Go implementation of the Model Context Protocol (MCP), enabling seamless integration between LLM applications and external data sources and tools.

21:18
strands-agents/harness-sdk harness-cli/v0.1.1 Sub-package +4 more

harness-cli/v0.1.1

06:17
anthropics/claude-agent-sdk-python v0.2.158 Patch

New Features

09:40
osaurus-ai/osaurus 0.25.12 Patch +1 more

Fall back from an unavailable or hung Foundation core model (#2861) by @tpae

15:48
2FastLabs/agent-squad typescript_1.1.5 Patch +1 more

Added `JevClassifier`: routes with TypeSafe's Jev decision model instead of an LLM. The registered agents become a typed `choice` question, and Jev returns the chosen agent plus a calibrated confidence, billed on input t

18:04
vllm-project/vllm-omni v0.30.0rc1 Minorpre

This release features 262 merged changes from 126 contributors, including 45 new contributors.

00:33
netease-youdao/LobsterAI 2026.9.23 Rolling

What’s Changed

22:29
mistralai/mistral-vibe v2.25.8 Patch

Minimal CLI coding agent by Mistral

23:29
callstack/agent-device v0.21.12 Patch

feat(ios): fold Duo headlessly through simulator HID by @thymikee in https://github.com/callstack/agent-device/pull/2762

03:19
Dao-AILab/flash-attention fa4-v4.0.0.beta32 Patchpre

[CuTe, SM100] Sparse MLA bwd: in-kernel recompute-P + token-chunked backward by @abcdabcd987 in https://github.com/Dao-AILab/flash-attention/pull/2816

16:21
anthropics/claude-agent-sdk-typescript v0.3.280 Patch

Added optional `fireReason` to the task-notification `SDKMessageOrigin`; a local host's declared scheduled-task fire is honored only in a process it started with `CLAUDE_CODE_HOST_SCHEDULED_RUN=1`

00:38
NVIDIA/TensorRT-LLM v1.3.0rc28 Patchpre

TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on…

14:34
lancedb/lancedb v0.40.0-beta.7 Patchpre +1 more

🎉 New Features

15:29
kserve/kserve v0.21.0-rc1 Patchpre

refactor(e2e): auto-assign llminferenceservice & consolidate markers by @vivekk16 in https://github.com/kserve/kserve/pull/5837

09:22
apache/maka v0.2.0-dev.47.20260922 Patchpre

Developer Snapshot 0.2.0-dev.47.20260922

03:17
microsoft/mcp Azure.Mcp.Server-3.0.0-beta.46 Patchpre IDE extension

_Improvements to Azure Backup and Storage Tools 🌐📚_

05:20