AI Briefing#open-source
The latest AI and developer news about #open-source, with the original source and a short summary.
The latest AI and developer news about #open-source, with the original source and a short summary.
The early release of the small model on Gemma 4 E4B scores 0.811 on the Werdykt benchmark and requires basal engine 2.8.0.
The release includes three model sizes ranging from 7.2M to 82M parameters, all supporting 24 kHz audio output and CPU-only execution via int8 ONNX weights.
The November 2026 update introduces elastic gateway scaling, HTTP-based worker task pulling, and a shared Slime-based training pipeline for mathematical reasoning.
The 26.6B parameter model supports seven languages and offers natural-language control over transcription style and context biasing.
llama.cpp has released v1.0.0, adopting a semantic versioning policy based on C API compatibility standards.
The free, MIT-licensed tool saves work to disk to survive context compaction and allows handoff between different AI coding agents.
Ace is an open-source tool that enables teams to work with AI agents in shared channels on their own machines, requiring Tailscale and Anthropic or OpenAI keys.
The stateless browser runs on Cloudflare Workers and supports Chrome DevTools Protocol for agent interaction.
The update reduces system prompts from ~8k to ~1k characters and uses plain text markers instead of JSON to ensure reliable performance on smaller local models like Gemma 12B.
The 4-bit version requires only 4 GB of VRAM and achieves approximately 13x real-time speed on an RTX 5090.
Gnomon is an open-source PostgreSQL extension that provides branch, diff, and merge capabilities for database data, allowing AI agents to safely modify production data with full auditability and undo functionality.
The release establishes strict semantic versioning rules for the C API and expands distribution channels to include Docker, Debian, and conda-forge.
Perplexity releases open-weight Decider v1.3 and Flash v1.3, with v1.3 scoring 69.0 and Flash 63.6 on out-of-domain extraction tasks, surpassing GLiNER large v2.5 (33.6).
The framework supports weight and activation quantization for LLMs, text-to-image, and text-to-video models while maintaining inference speed.
Paid APIs provide turnkey enterprise-grade speech synthesis with SLAs, while open source models offer data sovereignty and full architectural control at the cost of compute and engineering overhead.
The MIT-licensed tool uses encoder models and natural language inference to identify deviations and gaps between legal documents, avoiding the hallucinations and costs of generative LLMs.
nac uses a thread-and-episode architecture to keep AI agents aligned with original intent during extended experiments and infrastructure work.
The open-source plugin uses coding agents to find the fastest inference path for specific hardware configurations.
The new kernel leverages native matrix core shapes and SLM architecture, delivering significant speedups on ARL-H, PTL-H, and B70 Pro platforms.
The experimental preview models use reserved special tokens to separate instructions from data, achieving higher robustness against prompt injections without sacrificing general benchmark performance.
The team plans to release an anime model with alpha channels and multi-character references, alongside an editing model, while maintaining the Apache 2.0 license.
Open-source KQuant packs for Qwen models on Hugging Face achieve up to 3.7x lower KL divergence than public 4-bit baselines, with 6-15% faster inference on M5 Pro for 4B, 9B, and 27B models.
Tether AI released BotPsy-0.8B, based on Qwen3.5-0.8B, demonstrating superior tool-use performance compared to 2B models.
The free, MIT-licensed desktop app runs Whisper and Llama.cpp locally to generate meeting summaries without adding a bot to calls.