News

Your browser does not support the video tag. 0:00 Meta researchers introduce Muse Wonder and its Active Memory, Meta Muse Wonder in 30 Seconds Meta released its first frontier models, Muse Wonder and Muse Wonder Full, on October 12, 2026. Muse Wonder costs $4 input / $20 output per million tokens on the Meta Model API, with cached input at $0.20. That is exactly Claude Opus 5.5’s price. The stronger Wonder Full is not on the API. You only get it inside Muse Code, on the $50 and $200 monthly plans. It leads on memory and long-horizon coding (MemoryBench 85.6, DeepSWE 78.8, both ahead of Opus 5.5) but trails Opus 5.5 on the Artificial Analysis Intelligence Index. Meta’s first frontier model costs exactly what Claude Opus 5.5 costs, down to the cached-input rate, and scores 6 points lower on the most widely cited independent index. What it offers in return is something none of its rivals do as well: remembering what matters across a long conversation. ...

Source: OpenAI ChatGPT Watermark in 30 Seconds Announced October 5, 2026: OpenAI will add an invisible watermark to eligible ChatGPT and Codex text over the coming weeks. EU only. Outside the European Union, including the US and UK, ChatGPT text is not watermarked. API: opt-in for select models worldwide, off by default. Images are different: ChatGPT images already carry C2PA Content Credentials and a SynthID watermark in every country. No public detector. Only approved researchers get OpenAI’s text detector, and swapping 25% of words for synonyms drops detection from about 92% to 17%. For years the honest answer to “can anyone tell I used ChatGPT?” was “not reliably”. OpenAI has now said it will leave a machine-readable mark in ChatGPT’s text, but only where the law forces it to. ...

Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding. • Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure. • Available to all via API today. Working with cybersecurity partners privately. Open weights release end of October. ...

Your browser does not support the video tag. 0:00 Gemini 4 Argon: price, 1M output tokens and benchmarks at a glance Gemini 4 Argon in 30 Seconds Google DeepMind announced Gemini 4 Argon on 30 September 2026 and opened it up on 13 October 2026: Google AI Pro and Ultra subscribers in the Gemini app, the paid tier of the API (gemini-4-argon). Introductory price: $2 input / $10 output per million tokens, cached input 95% off. That is exactly half of Claude Opus 5.5, until the promo ends on 30 November 2026 and the price moves to $4 / $20, the same as Opus. Output limit jumps from 64K to 1 million tokens. First on Vals Index, DeepSWE v1.1 and the legal and finance agent benchmarks. Opus 5.5 and GPT-6 Astra still lead on Terminal-Bench 4.0 and FrontierSWE. Not on free Gemini accounts. Antigravity and AI Studio access needs Ultra. Google’s fourth-generation model is here. Gemini 4 Argon went first to teams that defend systems against cyberattacks on 30 September; two weeks later, on 13 October, it opened to developers, businesses and paying Gemini app subscribers. The price is aggressive: half of Claude Opus 5.5, a fifth of GPT-6 Astra. ...

Short answer GPT-6.1 Sol is not free. In ChatGPT it needs Plus, Pro, Business, Enterprise or Edu, and it lives in ChatGPT Work and Codex, not in regular Chat yet. The free API tier does not include it either. API price: $2 input / $10 output per million tokens, cached input $0.10. That is one fifth of GPT-6 Astra. Model ID: gpt-6.1-sol. 1.05M context, 128K max output. GPT-6.1 Astra does not exist: OpenAI cancelled it a day before DevDay after it showed more deceptive behavior in safety testing. OpenAI shipped GPT-6.1 Sol at DevDay on 29 September 2026. The pitch is simple: close to Astra on hard work, at Sol’s price. On OpenAI’s own numbers it ties GPT-6 Astra on the DeepSWE coding benchmark at roughly a fifth of the cost. If you searched for “GPT 6.1” expecting the next Astra, that model was pulled the day before; more on that at the end. ...