SupraLabs has released a curated chat title dataset with 115K samples, surpassing the previous record of 10K samples. The filtered dataset is available as `SupraLabs/chat-titles-filtered-115K`, while an unfiltered version with 150K samples is also provided, along with a legacy 12K dataset.
A local agent can access the web without paid APIs by using self-hosted SearXNG for search and Scrapling with Trafilatura for page extraction. The setup avoids vendor dependencies, uses open-source tools, and delivers search results and page content in Markdown format, with fallbacks for CAPTCHAs and security challenges.
SupraLabs has launched supra-title-FFT-preview, a chat title generation model trained on 115K samples from a filtered dataset, expanding coverage beyond its previous 12K-sample model. The model uses full fine-tuning on LiquidAI/LFM2.5-350M-Base with BF16 precision and is designed for single-purpose chat title generation, available via Hugging Face and supporting direct loading or vLLM deployment.
The ggml-webgpu project has added adapter toggles for half-precision (F16) support on Vulkan and NVIDIA GPUs. This update enables improved performance on compatible hardware across multiple platforms, including macOS, Linux, Android, Windows, and openEuler, with specific builds available for ARM and x64 architectures.
A discussion thread identifies the best local AI agents available today, emphasizing open-weight models and local hardware execution. The post defines 'agents' as autonomous software that self-determines actions without pre-programming, distinguishing them from tools like IFTTT or Apple Shortcuts, and sets rules requiring local deployment and open-source agent software as a primary focus.
LLaMA.cpp releases version b9729 with binaries for macOS, Linux, Android, Windows, and openEuler across multiple architectures. The release includes CPU, Vulkan, OpenVINO, SYCL, and ROCm support, along with a new UI package. Internal references to 'webui' have been removed.
A user asks how to integrate Gemma 4 12B with search capabilities using self-hosted AI models. They mention trying openwebui, which has issues with search engines like DDG, and seek alternatives that avoid using Brave or Google API keys.
LLaMA.cpp version b9728 introduces support for comment lines in --api-key-file configuration. The release includes pre-built binaries for macOS, Linux, Android, Windows, and openEuler across multiple architectures and hardware acceleration options, including Vulkan, CUDA, OpenVINO, and SYCL.
Users report that tools like LM Studio and Odysseus are limited by search engine request caps, often at 10 per day or hour, without API access. They suggest creating DuckDuckGo API accounts for better search access, but note that frontends rarely prompt for this. The post asks whether Hermes or Pi offer improved solutions.
The v0.17.6 release adds pre-approval tool input guardrails and SDK-only custom data for tool outputs. It also enforces a strict JSON-compatible contract for tool outputs and suppresses unnecessary whitespace warnings in tool names. @siddiksawani made their first contribution in this release.
v1.14.8a2 adds a single agent action to Flow definitions and validates CEL expressions at load time. It includes a new Datadog integration guide with an importable operations dashboard and updated snapshot and changelog for v1.14.8a1.
Datasette has released a new plugin, datasette-apps, enabling self-contained HTML+JavaScript applications to run in a secure iframe sandbox. These apps can execute read-only or write SQL queries against Datasette databases, with built-in security features like CSP headers and sandbox restrictions to prevent data exfiltration or unauthorized access.
Cohere Labs has released a 4-bit quantized version of North Mini Code on Hugging Face, reducing its size to approximately 20GB for local execution on devices like Macs. The model is now supported in Ollama, local runtimes based on llama.cpp, and via the OpenRouter API, improving accessibility for developers.
llama.cpp version b9703 includes a rework of the server's preset handling, removing remote HF preset support and deprecated functions. The release provides binaries for macOS, Linux, Android, Windows, and openEuler across multiple architectures and hardware acceleration options, including Vulkan, CUDA, OpenVINO, and SYCL.
mistral.rs v0.8.10 introduces OpenAI-compatible Agent Skills via a /v1/skills endpoint, enabling local models to execute domain-specific instructions and scripts without relying on frontier APIs. The update supports tools like file uploads and downloads via /v1/files and includes prebuilt binaries for Linux, macOS, and Windows.
CrewAI v1.14.8a introduces script and crew actions to FlowDefinition, adds DMN mode support, and enables flow execution without Python code. It also includes experimental support for JSON-first crews and ZIP deployment fallback, along with improved memory reset and token usage tracking.
llama.cpp has been updated to support downloading and managing models via API. This enables full model lifecycle control through API calls without requiring a user interface.
The Metal backend in llama.cpp has been extended to support f16 and bf16 tensor types for the concat operator, in addition to existing f32 and i32 support. This update includes specialized kernel templates, updated pipeline getters, and improved type-based kernel dispatch, with assistance from pi:llama.cpp/Qwen3.6-27B.
llama.cpp releases version b9688, adding model management and SSE realtime updates APIs. The release includes prebuilt binaries for macOS, Linux, Android, Windows, and openEuler, supporting various architectures and acceleration frameworks like Vulkan, CUDA, OpenVINO, and SYCL.
Gemma 4 E2B achieves 255 tokens per second in-browser on an M4 Max using WebGPU kernels. The demo and kernels are now available on Hugging Face for public use.