Tag
Fable 5.5 reportedly generated a complete animation in a single pass using only code. Chetaslua said that standard Opus 5.5 and Fable 5.1 produced noticeably worse results with the same prompt and provided a method for verification, but the report has not yet been officially confirmed.
Hume AI introduced the Hume Audio Decision API, a single-call model that takes audio plus a list of questions and returns answers with confidence scores, detecting tone, hesitation, laughter and speaker cues without needing a transcript.
Claude reportedly introduces a new $200 Ultra plan that removes the five-hour usage limit, though new accounts or users from different regions may face verification checks.
Rumors suggested Qwen 4 was expected to be released at the start of the month, but it has not appeared yet, prompting questions about its release timing.
WhatsApp's official account congratulated Kimi on making history, likely marking a notable achievement for the Kimi AI model.
Nous Research hints at an impressive new AI model release with a brief hype post, though no concrete details are provided.
Astra reportedly outperforms all other models on robot intelligence benchmarks, marking a notable advance in embodied AI capabilities.
Pika H3 Flex 是一款新发布的开源权重视频生成模型，主打专业迭代工作流，可在 7 秒内生成 15 秒带音轨的视频，支持导出音频分轨、最高 4K 分辨率草稿模式，以及对角色、声音和运镜的后期编辑。
Shubham Saboo introduces Gemini 4 Carbon, a new model focused on long-horizon tasks, released just two weeks after Gemini 4 Argon.
Google's new Carbon model debuts at the #1 spot on the Artificial Analysis leaderboard with a score of 59, marking a notable top ranking for the company.
Google DeepMind teases or announces 'Gemini 4 Carbon', the next generation of its Gemini model family.
Nace 发布 Drex Max 1.1，这是一个在分布外泛化上显著提升的多模态决策模型，拥有 262K 上下文窗口，并在 Decision Index 上取得 70.8 的成绩。
A tweet speculating on the significance of a purported Claude Opus 5.5 release, with claims that if its results hold up it could mark the biggest single day in the history of math and theoretical computer science.
GPT-6.1 Terra 发布，OpenAI 称其在快速、能干的计算机使用（computer use）方面实现新突破，能理解屏幕内容、跨应用执行操作，并以更低价格提供更高效的工作能力。
CodeRabbit is using OpenAI's GPT-6.1 Sol as the primary model in its security agent to strengthen code security across complex codebases, enabling cross-file risk analysis, dependency investigation, exploit verification in sandboxes, and automated fix pull requests.
OpenAI 发布 GPT-6.1 Terra，一款兼顾视觉推理、计算机操作与自然写作的高效模型，API 价格较 GPT-5.6 Terra 低 50%，在 Agents' Last Exam、MMMU-Pro、OSWorld 2.1 和 Terminal-Bench 4.0 等基准上大幅提升，同时显著降低幻觉与事实错误率。
A humorous post noting how quickly LinkedIn engagement farmers latched onto eMudhra's 0x4 release, poking fun at the hype cycle around new AI models.
SoMark released OXR, a 1.3B open-source document parsing model that tops olmOCR-Bench, OmniDocBench and DocParsingBench while delivering 1.60× DeepSeek-OCR-2's throughput on a single RTX PRO 6000. It handles footnotes, formulas, tables and handwriting across 13 element types with coordinates, powered by RSI.
Researchers from Mass General Brigham report an AI foundation model trained on over 40 million facial photographs that can predict cancer patients' mortality risk, building on their earlier FaceAge system for estimating biological age from photos.
MiniMax 发布 M3.1 Flash 新模型，在复杂 agentic 任务、计算机使用、软件开发与网络安全等多类基准上全面提升，其中 DeepSWE v1.1 达到 73.7%，跻身顶尖模型行列。