Daily AI news, sourcedUpdates / Models
Updates Models · 505 items · newest issue first · page 9 of 26
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- torch-pointcloud puts pretrained 3D point cloud models behind one timm-style create_model() factoryUnconfirmedr/MachineLearningOct 14Models
- Muse Wonder (high) scores 93.4 on the Extended NYT Connections benchmark, ahead of Muse Spark 1.3 at 85.1Unconfirmedr/singularityOct 14Models
- Meta says Muse Spark open weights are coming soon but Muse Wonder will not be open-sourcedUnconfirmedr/LocalLLaMAOct 14Models
- Grok 4.7 is now available on Amazon Bedrock
- Developer converts Qwen2.5-1.5B into a Mixture-of-Experts model and releases GGUF builds of the 0.5B and 1.5B conversionsUnconfirmedr/LocalLLaMAOct 13Models
- LocalLLaMA user runs the 'horsey challenge' on Qwen 3.8 Flash-Next in NVFP4 with the strata-nvfp4 forkUnconfirmedr/LocalLLaMAOct 13Models
- mlx-community publishes MXFP8 and MXFP4 MLX builds of MiniCPM-V 4.7 1B and 35B-A3B for Apple SiliconUnconfirmedHugging Face @mlx-communityOct 13Models
- Blind quiz: 60% of respondents judged a Claude Opus 5.5 oboe concerto human-written, ahead of Handel and VivaldiUnconfirmedX @aug5thmusicOct 13Models
- User splits a 27B ternary model across two Android phones over USB, reporting about 2.8 tok/sUnconfirmedr/LocalLLaMAOct 13Models
- LocalLLaMA user starts a Qwen 3.8 Flash Next Q3_S benchmark after strong Exo results in OpenCodeUnconfirmedr/LocalLLaMAOct 13Models
- Kling AI features PROMPTR Studios' Simon Meyer on making a viral music video with Kling 4.0
- Meta's Alexandr Wang shares an early tester's claim that Muse Wonder matched Astra, Fable and Opus on their benchmarksStatementX @alexandr_wangOct 13Models
- Intercom's Fin Apex 2.0 customer service models are fine-tuned from NVIDIA Nemotron 3 Super and NanoOfficialX @NVIDIAAIOct 13Models
- Artificial Analysis launches a FrontierCode 1.1 leaderboard, with Claude Opus 5.5 on top at 52%OfficialX @cognitionOct 13Models
- Runway adds Ideogram 4.5 to its developer API for text-to-image with legible typography
- User reports running a 4-bit Qwen3.8-27B with 262k context and vision on a 16GB GPU at 80 tokens per secondUnconfirmedr/LocalLLaMAOct 13Models
- Meta launches Muse Wonder, its first frontier model, on the Model API and Muse Code, with Muse Wonder Full for subscribersOfficialMetaOct 13Models
- Security engineer Micah Lee urges people to drop OpenAI and Anthropic in favor of open-weight modelsStatementmicahflee.comOct 13Models
- OpenBMB publishes MiniCPM-V 4.7 in 1B and 35B-A3B sizes, the first MoE model in the on-device vision series
- Amazon Bedrock offers OpenAI GPT-6 models through global cross-Region inference from Thailand, Malaysia, Singapore, Indonesia and TaiwanOfficialAWSOct 13Models