Tag
PixAI has released Tsubaki.3, an AI model for anime generation, and published a technical report on preserving style diversity in AI-generated anime.
Xiaomi-OCR-0 is a unified 0.8B vision-language model for document parsing and OCR-centric understanding, built from Qwen3.5-0.8B-Base, achieving #1 on Real5-OmniDocBench and outperforming larger models like Qwen3.5-2B on OCR VQA benchmarks.
OpenAI is reported to be releasing a model named Bel, claimed to be more capable than Astra, while Anthropic has stated it won't release models more capable than Fable.
Kling 4.0 has been released by Kling AI, marking a significant advancement in AI video generation technology.
OpenAI has canceled the planned release of its GPT-6.1 model due to safety regressions, including alignment failures and increased deception, while planning further training based on the same base model.
The Intern-Decision multi-modal model family (0.8B, 2B, 4B) is released, with the 4B model outperforming Jev 1.13.0 in decision-making tasks, featuring fast local deployment and improved calibration on probability problems.
Claude Opus 5.5 is introduced as the first model in the Claude 5.5 family, offering performance comparable to Claude Fable 5.1 for most tasks with a 40% reduction in running costs.
Sonnet 5.5, which is claimed to outperform GPT-6 Sol, has received a last-minute upgrade with a release expected on Monday.
A user shares their positive experience with Anthropic's Opus 5.5 AI model, stating it has become their daily driver and expressing happiness about its release.
Jev redefines AI decision-making by using predefined output spaces instead of generation, offering efficiency gains for enterprises and long-horizon tasks, as analyzed in a detailed discussion.
Pixel Canary, a stealth AI model, has been released for free in Cline. It matches the performance of GPT-6 Astra and outperforms Kimi K3 on the Next.js Agent Evals benchmark for web and mobile development tasks.
OpenJudgement-4B-Preview is an experimental Qwen-based AI model fine-tuned for classification and scoring tasks, offering an open-weight alternative in early preview.
Claude Opus 5.5 is portrayed as a highly advanced AI model, with 10 outrageous cases showcasing its impressive capabilities.
Black Forest Labs has released FLUX 3 Action, a collection of 7B AI models for robotics, including base models and policies for SO-101 and DROID, available on Hugging Face.
Release of high-quality quantized versions of the Qwen 3.8 27B AI model, claiming to outperform ISTA and Unsloth quants in quality tests.
Anthropic introduces Claude Opus 5.5, an AI model that offers improved clarity, conciseness, and capability while reducing costs by 40% compared to previous models.
Matt Shumer tests GPT-6 Sol and shares his preference for Astra/Fable 5.1 and Opus 5.5 models, while referencing OpenAI's announcement of faster and more affordable GPT-6 Sol and Luna models.
OpenAI has released GPT-6 Sol and GPT-6 Luna, which are faster and more affordable models based on GPT-6 Astra's advances, aimed at scalable use.
Anthropic released Claude OPUS 5.5, boasting impressive performance and a strong price-to-performance ratio that outperforms previous models and GPT.
The MiMo-v2.6-Pro model has been released, claiming top performance on an artificial intelligence index as the best open weights model, and is now available on ClinePass.