https://topaiproduct.com/2026/08/29/cohere-parse-5-loses-the-benchmark-wins-at-1-50-per-1000-pages/Top AI Producten2026-08-29T16:08:40+00:00Cohere Parse 5 loses the benchmark, wins at $1.50 per 1,000 pageshttps://topaiproduct.com/2026/08/29/google-expert-intelligence-gemini-notebook-x-play-books-turns-100000-paid-books-into-ai-sources/Top AI Producten2026-08-29T16:06:05+00:00Google Expert Intelligence (Gemini Notebook × Play Books) turns 100,000 paid books into AI sourceshttps://topaiproduct.com/2026/08/29/tencent-hy4-preview%ef%bc%88770b-%e5%bc%80%e6%ba%90-moe%ef%bc%8c1m-%e4%b8%8a%e4%b8%8b%e6%96%87%ef%bc%89scores-64-3-on-deepswe-up-from-28-a-generation-ago/Top AI Producten2026-08-29T10:07:44+00:00Tencent Hy4 preview（770B 开源 MoE，1M 上下文）scores 64.3 on DeepSWE — up from 28 a generation agohttps://topaiproduct.com/2026/08/29/claudeforce-salesforce-makes-claude-its-default-brain-and-moves-its-whole-crm-inside-claude/Top AI Producten2026-08-29T10:05:25+00:00Claudeforce: Salesforce makes Claude its default brain — and moves its whole CRM inside Claudehttps://topaiproduct.com/2026/08/28/115000-stars-in-11-weeks-ponytail-the-skill-that-makes-your-ai-agent-code-like-the-laziest-senior-dev/Top AI Producten2026-08-29T04:08:27+00:00115,000 stars in 11 weeks: ponytail — the skill that makes your AI agent code like the laziest senior devhttps://topaiproduct.com/2026/08/28/openai-cuts-off-cursor-model-access-winds-down-after-spacexs-60b-acquisition/Top AI Producten2026-08-29T04:06:06+00:00OpenAI cuts off Cursor: model access winds down after SpaceX’s $60B acquisitionhttps://topaiproduct.com/2026/08/28/openmontage-hits-53240-github-stars-your-coding-agent-is-now-a-video-studio/Top AI Producten2026-08-28T22:07:36+00:00OpenMontage hits 53,240 GitHub stars: your coding agent is now a video studiohttps://topaiproduct.com/2026/08/28/archify-adds-4561-github-stars-in-one-day-the-agent-skill-that-stops-your-coding-agent-from-hand-drawing-diagrams/Top AI Producten2026-08-28T22:05:05+00:00archify adds 4,561 GitHub stars in one day — the agent skill that stops your coding agent from hand-drawing diagramshttps://topaiproduct.com/2026/08/28/hugging-face-launches-microduck-a-399-open-source-bipedal-robot-you-train-yourself/Top AI Producten2026-08-28T16:08:27+00:00Hugging Face launches Microduck, a $399 open-source bipedal robot you train yourselfhttps://topaiproduct.com/2026/08/28/us-judge-rules-pentagon-blacklisting-of-anthropic-unlawful-the-first-courtroom-win-for-an-ai-safety-red-line/Top AI Producten2026-08-28T16:05:21+00:00US judge rules Pentagon blacklisting of Anthropic unlawful — the first courtroom win for an AI safety red linehttps://topaiproduct.com/2026/08/28/spline-v2-rewrites-the-entire-3d-editor-then-hands-the-keys-to-claude-code-and-cursor/Top AI Producten2026-08-28T10:07:59+00:00Spline V2 rewrites the entire 3D editor — then hands the keys to Claude Code and Cursorhttps://topaiproduct.com/2026/08/28/claude-cowork-built-in-browser-anthropic-lands-two-weeks-after-openai-killed-atlas/Top AI Producten2026-08-28T10:05:17+00:00Claude Cowork Built-In Browser (Anthropic) lands two weeks after OpenAI killed Atlashttps://topaiproduct.com/2026/08/27/gemini-omni-1-1-flash-hits-ga-google-prices-4k-ai-video-at-0-30-a-second/Top AI Producten2026-08-28T04:07:24+00:00Gemini Omni 1.1 Flash hits GA: Google prices 4K AI video at $0.30 a secondhttps://topaiproduct.com/2026/08/27/gemini-3-5-transcribe-google-hits-2-6-word-error-rate-and-deletes-your-ums-for-you/Top AI Producten2026-08-28T04:05:24+00:00Gemini 3.5 Transcribe (Google) hits 2.6% word error rate — and deletes your “ums” for youhttps://topaiproduct.com/2026/08/27/glm-5-3-flash-z-ai-is-ox-alpha-the-mystery-model-that-hit-2-on-openrouter-running-on-chinese-chips/Top AI Producten2026-08-27T22:07:21+00:00GLM-5.3-Flash (Z.ai) is Ox Alpha — the mystery model that hit 2 on OpenRouter running on Chinese chipshttps://topaiproduct.com/2026/08/27/anthropic-model-hardware-standard-mhs-compresses-weeks-of-lab-hardware-integration-into-hours/Top AI Producten2026-08-27T22:04:48+00:00Anthropic Model Hardware Standard (MHS) compresses weeks of lab hardware integration into hours