Инструменты ИИ и обновления
Meta released Muse Spark Open on October 15, 2026: an 816 billion parameter open-weight model built for coding and agentic work. It leads JobBench and MCP-Atlas, trails Chinese rivals on pure coding benchmarks, and runs on two Mac Studios.
Инструменты ИИ и обновления
xAI released Grok Voice 3.0 on October 14, 2026. It leads the Artificial Analysis Speech-to-Speech Index at 86.9, costs $0.08 per minute, and answers in 0.81 seconds. Here is how it compares to OpenAI GPT-Live-1 and Google Gemini Live.
Инструменты ИИ и обновления
Meta released Muse Wonder and Muse Wonder Full on October 12, 2026. Here is how they score, what they cost, and whether Active Memory is a real advantage.
Новости ИИ
Nvidia is reportedly in talks to acquire or invest deeper in Reflection AI, the open-weight lab behind the Beam model, valued at $25B.
Инструменты ИИ и обновления
Claude Haiku 5.5 costs $0.10 per 1M input tokens, matches GPT-6 Luna's price, and beats it on benchmarks. Pricing, context, speed, and which to pick, compared.
Новости ИИ
TypeSafe AI raised $870 million at a $7.5 billion valuation for Jev, a model that returns decisions instead of text. How it works, what it costs, and where OpenAI and Amazon's rivals fit in.
Новости ИИ
Google unveiled the Googlebook on September 21, 2026: a premium Android-based laptop line from Acer, ASUS, Dell, HP and Lenovo with deep Gemini integration. Here is everything it is, and who should buy one.
Инструменты ИИ и обновления
Google Cloud just launched the Gemini agent: one AI worker that plans tasks, routes between Gemini and Claude models, and pauses when spend caps hit. We break down what it does, what it costs, and who should care.
Инструменты ИИ и обновления
Модель Mistral с 1,05 триллионами параметров, прозванная Le Chonk, представлена в публичном предварительном просмотре. Мы сравниваем её с GPT-6 Luna и DeepSeek V4.1 Flash на бенчмарках, разбираем цены и объясняем выпуск открытых весов в конце октября.
Инструменты ИИ и обновления
Google's Nano Banana 2.1 leaked via the Flow build. We compare it with GPT Image 2.5 and Nano Banana Pro on quality, speed, and price.
Инструменты ИИ и обновления
Reflection AI's Beam is a 501B open-weight model that claims GLM-5.2 level reasoning at 3-4x less compute. The specs, the company-published benchmarks, and the AI factory strategy, explained.
Инструменты ИИ и обновления
Tibo's "6.1 coming soon" post pointed at Ultrafast, OpenAI's paid speed tier: 300 tokens/sec at 6x the price. The numbers, the NVIDIA twist, and the math.