ERNIE 5.5 Flash: Baidu's 280B Open MoE, Specs & Benchmarks
ERNIE 5.5 Flash Base is Baidu's open 280B MoE with 14B active, 33.5T tokens and 1M context. Specs, benchmarks vs DeepSeek V4 Flash, architecture and how to run it.
Sumus 博客
关于 AI 队友的使用指南、工作方法与产品新进展。
Claude Personal Finance: How It Works, Safety & vs ChatGPT
Claude Personal Finance connects your bank and cards via Plaid, read-only, for US Pro, Max and Ultra users. What Claude sees, how safe it is and how it compares with ChatGPT Finances.
GPT-6 Cyber: Benchmarks, Daybreak Access & vs GPT-6 Astra
GPT-6 Cyber is OpenAI's security model: 48.5% on ExploitGym, 81 new Firefox bugs, Critical cyber rating. Benchmarks, how to get Daybreak access and how it differs from Astra.
Perplexity Teammates: Price, Features & a Cheaper Alternative
Perplexity Teammates are always-on work agents on the $200 Max plan, with Slack and Computer integration. What they do, vs OpenAI Dots and Claude Duet, and a free alternative.
Ling 3.2 Flash: Ant's 560B Multimodal Model, Price & Benchmarks
Ling 3.2 Flash is Ant Ling's native multimodal MoE: 560B total, 25B active, 1M context, $0.30/$0.90 and free for two weeks. Benchmarks vs Opus 5 and how to try it.
Tencent Hy4 (Hunyuan 4): Price, Benchmarks & What's New
Tencent Hy4, the full Hunyuan 4 release, adds vision and computer use at $0.90/$2.70 per 1M tokens. Specs, benchmarks, Hy4 vs Hy4 Preview and where to try it.
Gemini Omni 1.2 Flash: 15s Clips, $0.07/s & vs Seedance 2.5
Gemini Omni 1.2 Flash makes 15-second AI videos at $0.07 per second, 30% cheaper than 1.1. What's new, pricing, where to use it and how it compares with Seedance 2.5.
ChatGPT GIF: How to Make GIFs in ChatGPT (Images 2.6)
ChatGPT GIF making is built in with Images 2.6: turn a prompt or photo into a looping GIF, change speed, edit frames. Steps, prompts that work and limits.
GPT Image 2.6: Flare vs Sunburst, GIFs, Speed & Price
GPT Image 2.6 (ChatGPT Images 2.6) launched Nov 4, 2026: GIFs, reference images, 50% faster, Flare in 6 seconds, Sunburst for quality, SVG export and API price.
Ripple 65B: Specs, Benchmarks, License & How to Run It
Ripple-65B Preview by Ricursive is a 65B dense open-weight model (Apache 2.0) scoring 80.5 on Terminal-Bench 2.1. Specs, benchmarks, access and how to run it.
Crouton Alpha: Who Made the Free 1M-Context Stealth Model?
Crouton Alpha is a free stealth model on OpenRouter with 1M context and text, image, audio and video input. What's confirmed, who may be behind it, how to try it.
LongCat 2.5: Meituan's AI Model, Benchmarks, Price & API
LongCat 2.5 is Meituan's 1.6T-parameter multimodal MoE for coding agents: 48B active, 1M context, $0.30/$1.20 API. Benchmarks, setup in Claude Code, limits.