AI releases
25 itemsEvery AI release and brief surfaced by Freedom.Tech.
llama.cpp b9788
Sycl tp stage1; SYCL: tensor parallelism (--split-mode tensor) for dual-GPU; Small (nelem < 32768): FP32 direct memcpy + per-device ADD kernel chained via depends on(memcpy event). 4 SYCL submissions/call; Fix comments
Mistral Common 1.11.4
Add chat templates integration; Deprecate RawAudio in favor of str | bytes; Generalize normalizer aggregation with dual separators; Consolidate multimodal ContentChunk support for all message roles
HelixML 2.11.28
Host docker-compose stacks + persist certmagic in filestore in https://github.com/helixml/helix/pull/2718; Filter org-chart and non-external agents from spec task agent switchers in https://github.com/helixml/helix/pull/2717; Self-
Ollama 0.30.11
Launch: add thinking capability detection to opencode in https://github.com/ollama/ollama/pull/15434; Launch: auto-install Claude Code in https://github.com/ollama/ollama/pull/16802; Launch: auto-install opencode when missing in https://github.com/ollama/ollama/pu
openclaw 2026.6.11-beta.1
Richer operator workflows: openclaw agent --message-file and the RAFT CLI wake bridge add practical file-driven and remote wake-up paths. ( 93351, 95497) Thanks @ooiuuii and @vincentkoc; Safer plugin distribution: additional official plugins are externalized cleanly, with bundled plugin icon metadat
Claude Agent Sdk Typescript 0.3.191
Added old source field to NotebookEdit tool results for replace and delete operations, enabling inline diffs; Added seven day overage included to SDKRateLimitInfo.rateLimitType for per-model weekly usage limits; Added model scoped array to usage response for per-model weekly limit windows with utili
LlamaIndex 0.14.23
[2026-06-24]; Llama-index-callbacks-argilla [0.5.0]; Llama-index-core [0.14.23]; Multimodal synthesis part 2 (#21561)
Elevenlabs Android 0.10.0
Expose reconciled message transcript as StateFlow; Deprecate text/agent callbacks in favor of event-id variants; Add Android release command
Qwen Code 0.19.2
Add remote LSP status route; Cap fork turns and bubble fork permission prompts; Transcribe images to text for text-only models
Jan 0.8.3
Merge release/v0.8.2 to main in https://github.com/janhq/jan/pull/8250; Use Windows copy command in b
LocalAI 4.5.0
The redesigned Home: console with a built-in assistant and chat
ComfyUI 0.26.0
Partner Nodes; Add the checkbox to disable head drawing in n
Client Ts typescript 2.3.0
OpenAPI Doc; Speakeasy CLI 1.763.6 (2.884.13) https://github.com/speakeasy-api/speakeasy; Typescript v2.3.0; NPM v2.3.0 https://www.npmjs.com/package/@mistralai/mistralai/v/2.3.0 -
Elevenlabs Python 2.54.0
Allow passing headers, follow redirects from ElevenLabs, AsyncElevenLabs to parent/base class; @pchng made their first contribution in 797
Kimi Code 0.19.1
Fix ACP editors such as Zed failing to start a new thread
langchain== 1.3.11
Release(langchain): 1.3.11 ( 38377); Only set strict=True on tools for OpenAI-compatible models in ProviderStrategy ( 38370)
langchain-openai== 1.3.3
Release(openai): 1.3.3 ( 38375); Drop response item ids when store is false ( 38372); Only set strict=True on tools for OpenAI-compatible models in ProviderStrategy ( 38370); Drop stop from Resp
Kimi Cli 1.48.0
Round-trip empty reasoning content; Escalate repeated-tool-call reminders and force-stop on dead-end streak; @sailist made their first contribution in 2467
Cline CLI 3.0.29
Costs are now hidden for Cline free models; Fixed Z.ai model metadata resolution for Z.ai models accessed through the Cline provider; Reverted the model-name-only display change from v3.0.28; The model picker, selector, and status bar return to their previous display behavior
Hermes Agent 0.17.0
Release Date: June 19, 2026 Since v0.16.0: 1,475 commits 800 merged PRs 1,693 files changed 235,390 insertions 50,730 deletions 300+ issues closed 245 community
whisper.cpp 1.9.1
Release: v1.9.1 in https://github.com/ggml-org/whisper.cpp/pull/3892
ComfyUI 0.25.1
Partner Nodes; Add support for Kling V3-Turbo model (https://github.com/Comfy-Org/ComfyUI/pull/14528)
Z Ai Sdk Java 0.3.5
Update tts asr api and samples; Support async generate images
Hugging Face Transformers Patch 5.12.1
Fix peft lower bound 46605; Mistral common backend fix 46667
vLLM 0.23.0
Gemma 4: Added encoder-free Gemma 4 Unified support ( 44429) and Gemma 4 MTP ( 43241), plus numerous accuracy and startup fixes; Transformers v5 compatibility: vLLM now targets Transformers v5, with vendored MiniCPM-V/O processors ( 44282) and compatibility fixes for Sarvam ( 38804) and Voxtral ( 44