全部 AI 动态

Important note regarding Grok @Bot: Going forward, @SpaceX will use the best back end model for any given task, including Claude Opus 5.5, MidJourney, Suno and other leading APIs. Whatever is most likely to give you the best outcome.

认识一下 Mistral Large 4，又名 Le Chonk。 • 1T 参数，原生多模态。49B 激活。 在综合基准测试中，它是来自美国或欧洲的最佳开放权重模型。 • 在关键工作负载上达到最先进水平，包括网络防御、制造业和金融，并且在视觉定位方面超越了闭源前沿模型。 • 端到端在欧洲打造，并可通过我们自己的 Mistral Cloud 基础设施从欧洲部署。 • 今天即可通过 API 向所有人开放。正与网络安全合作伙伴私下合作。 开放权重版本将于十月底发布。

Pulse by @smallest_AI currently ranks #1 in the Diarization + ASR track by DER, as the newest entry on the Voice Arena Diarization Bench. The Diarization + ASR track ranks Speaker-Attributed ASR systems that return speaker labels together with the transcript. Seven systems are in it today. Pulse scores 24.4% DER at a strict 0 ms collar. The next system, ElevenLabs Scribe v2, scores 40.7%. That is 1.7x lower error than the next best in the track. Most of the gap is missed speech. For every other system in the track, missed speech is the largest part of the error, above 30% for each of them. Pulse misses 4.4% of reference speech, the lowest in the track. On the full board of 13 systems, dedicated diarization models and Speaker-Attributed ASR systems together, Pulse places 5th. It is the highest-ranked Speaker-Attributed ASR system on the board and sits ahead of two dedicated diarization models. The result holds across settings: In-person recordings : 24.7% DER In online calls : 22.4% DER Congratulations to @smallest_AI 👏 View full results, per track and per condition: http://voicearena.com/diarization-bench

Nano Banana 2.1 is here! This upgraded version outperforms our previous models across the board, with notable leaps in: - visual design - mask-based editing - subject consistency - more natural-looking imagery Try it now in @GeminiApp, @GoogleAIStudio, and more.

You can now control agents on your computer from your phone. Check in, reply, or start new tasks from the Cursor iOS app.

Bloomberg: SpaceX is in early talks to borrow $40B for Nvidia chips, with Apollo Global Management leading. The package would pair about $10B in bank loans with $30B of investment-grade debt and should close in 2027, according to the Financial Times. Musk has earlier said SpaceX will build exclusively on Nvidia hardware. Many of those chips get rented out, because SpaceX now runs a cloud business in which Anthropic could pay it up to $84.5B through 2029.

We’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. https://github.com/openai/math

OpenAI published the full papers behind its open-problems claim: 722 manuscripts covering 372 math results. OpenAI built this catalogue by giving its internal model roughly 4K research problems during testing, then grouping and filtering the output into 372 significant result families. Almost all of those results came from the same standard setup, and each used on average the computing equivalent of 3 hours of ChatGPT Pro thinking.

We’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. https://github.com/openai/math

让你的应用通过 Decisions API 近乎实时地选择正确的模型、工具或操作，现已面向所有开发者开放公开测试版。 Decisions API 通过 Responses API 做出决策的速度比 GPT-6 Luna 快达 10 倍。

We’re expanding our Cyber Verification Program to give security professionals broader access to our most capable models. Through this program, verified security professionals can access Claude Mythos 5.1, Opus 5.5, and Sonnet 5.5 with safeguards designed for defensive work. We’re also opening up new tiers to allow for authorized offensive work, like penetration testing and red-teaming. https://www.anthropic.com/news/cyber-verification-program

https://x.com/i/article/2107473094155927552

AutomationBench Verified is out. We went through all 600 public tasks in @Zapier's AutomationBench and checked every verifier. • Agents flagged 323 verifiers as suspicious, human review confirmed 206 real bugs, all 206 are fixed. • We replayed 1,235 Kimi K3 runs on the old and fixed verifiers and 344 of them (27.9%) got a different grade • Where verifiers were too strict, pass rate went from 18.8% to 43.8% • Where they were too lenient, it dropped from 60.2% to 49.7% Audit and dataset links below.