Tag
Audio8's open-source TTS model reached #1 on Hugging Face's TTS trending list and #10 overall, with the team expressing gratitude and plans to keep improving.
Alibaba Qwen announces that Qwen3.8 Max is now available in OpenCode Go.
SupraLabs releases SupraBrain-50M, a hybrid language model combining Gated DeltaNet, Sliding-Window Attention, and Surprise-Gated mechanisms, achieving near-parity with Supra-Base-50M despite training on far fewer tokens.
The tweet mentions that GLM-5.3 has been launched on GLM's official website, and marvels at the pressure Deepseek v4 flash and Qwen 3.8 max are facing.
MiniMax has released an open-weight video model, though with regional restrictions in the license.
Alibaba Qwen announces that open weights for Qwen3.8-Max (2.4T parameters) and Qwen3.8-27B will be released next week.
Qwen announced Qwen3.8, including a new 27B model, generating excitement for local deployment.
Alibaba officially released Qwen3.8, with 2.4 trillion parameters and 1M context. Coding and office capabilities have been significantly improved, entering the global top tier. The author conducted a hardcore real-world test using a single-file HTML N-body simulation.
A tweet jokes that while Anthropic is cautious about releasing powerful AI, Qwen has released Qwen3.8-Max, a new model for coding and cowork, asking users to star their repo.
阿里云Qwen宣布下周开源Qwen3.8-Max和Qwen3.8-27B的权重，其中Max为2.4T参数模型，主打编码与协作。
Hugging Face repository providing GGUF quantizations of MiniMax-H3 models for use with ComfyUI, including directory structure and links to required VAEs.
Qwen 3.8-Max official version released. The video demonstrates the Qwen Office and Qwen Code tools, focusing on programming and collaborative office capabilities.
Alibaba announces Qwen3.8-27B open-weights release, capable of running locally on 17GB RAM/VRAM, alongside the larger Qwen3.8-Max.
Qwen3.8-Max sets a new benchmark for coding and collaborative work capabilities in AI models.
Qwen 3.8-Max, a 2.4 trillion parameter model, is now available with open weights coming next week, delivering improvements in coding, work, research, and long-horizon tasks.
Poolside released an updated checkpoint for the Laguna-S-2.1-NVFP4 model on Hugging Face, noting that weights have changed and earlier downloads should be re-fetched.
SyzygyResearch releases Mach-1-Additive-35B, a quantized 35B model achieving 95% mean retention across 12 benchmarks relative to its full-precision teacher, outperforming other quantization baselines. The card includes per-benchmark retention scores, speed, and teacher model comparisons.
OpenHands announces the release of DeepSeek-v4-Flash, offering it for free to OpenHands Cloud users for a limited time by selecting 'openhands/deepseek-v4-flash' as the LLM.
DeepSeek released DeepSeek-V4-Flash-0731, a 304B parameter model with enhanced agentic capabilities, priced at $0.14/M input and $0.27/M output, punching above its weight and ranking as the best value-per-intelligence model according to Artificial Analysis.
Nous Research has opened free access to FLUX 3 Preview for all Nous Portal users, including free tier, and is running a short film contest with prizes.