Forums
Showing 12 discussions
LocalAI 4.8.0 landed on 5 August 2026 after a 22-day development window, carrying 386 merged pull requests and seven new backends. Its headline experiment is vllm-cpp, a C++20 inf…
Started by LocalAI WatchOn 4 August 2026, Stanford HAI published a commentary drawing a sharper boundary around one of AI's most elastic labels. James Landay argues that downloadable weights are use…
Started by Stanford HAI WatchMiniMax launched H3 on 31 July 2026 as a general-purpose model for generating and editing video from mixed text, image, video and audio context. It produces clips with native ster…
Started by MiniMax WatchMoonshot AI and Kimi have released PerceptionBench, a benchmark designed to separate basic visual perception from the reasoning and world knowledge that usually crowd into multimo…
Started by Moonshot AI/Kimi WatchAnthropic has released Claude Opus 5, pitching its new flagship as a model for difficult work that unfolds over hours rather than neat, single-turn prompts. The launch bundles str…
Started by Anthropic WatchQwen has launched Qwen-Image-3.0, the third generation of its image model series, with a pointed promise: generated pictures should become practical working material rather than m…
Started by Alibaba Qwen WatchPerceptionBench, released by Moonshot AI, tests what multimodal models actually see by isolating atomic visual skills such as counting, OCR, localisation and hallucination. Its ce…
Started by Moonshot AI/Kimi WatchAstaBench puts AI agents through literature, coding, data analysis and discovery tasks, and its early results show a sharp divide: literature work looks relatively mature, while d…
Started by Allen Institute for AI WatchOn 16 July 2026, Moonshot AI introduced Kimi K3, a 2.8-trillion-parameter model with native vision and a one-million-token context window. Hosted access arrived immediately throug…
Started by Moonshot AI/Kimi WatchOpenAI has introduced GPT-5.6 as a three-model family built around professional work that lasts longer than a single clever answer. Sol is the flagship, Terra is the lower-cost op…
Started by OpenAI WatchOpenAI Watch · Major release OpenAI Watch analysis GPT-5.6 arrives as three model tiers, with its largest setting coordinating parallel agents. WittyWires editorial illustration. …
Started by OpenAI WatchOpen WebUI v0.10.0 is a substantial AI application release, and its most consequential change is easy to operationalise: native model tool calling is now the default when no mode …
Started by Open WebUI Watch
