Hy4 is a big step up from Hy4 preview in coding & agentic benchmarks, from 26.3 to 44.4…
This is a dev post classified by AGK Intelligence as AI dev tools (a launch), kept by the Dev Radar because it carries real work, not commentary.
Hy4 is a big step up from Hy4 preview in coding & agentic benchmarks, from 26.3 to 44.4 in terminal bench 4. It now also has vision and computer use 🔥 Non-nerfed Opus 5.5 gets 66.4 on terminal bench. Will share results soon!
Posted by Mia (45k followers) 1 h ago · 43 likes · 5.3k views · view the original post on X. Kept by the Dev Radar as AI dev tools.
More dev work like this
- HY4 just got vision. — @Chinazhidx
- Tencent released Hy4, its new flagship model with performance among the top open models… — @testingcatalog
- 照着一个网站做练习，截图发给 AI，电脑上看着差不多，缩到手机宽度，间距、换行和菜单又全对不上。 — @GitHub_Daily
- Running coding agents in parallel shouldn’t mean losing track of their changes. — @DanKornas
- 百度开源了一个 OCR 模型，光上个月在 HuggingFace 上就被下载了 165 万次，名字却没多少人听过。 — @bkdgiffug
- Muse Code is coming shortly too — @robinebers
- 混元4来了，从评分上和kimi-k3.1不相上下？ — @mylifcc
- DeepSeek STQ 1.75bpw v2.0 is here! Custom tunings cut ~15% errors in every category,… — @Beamsters1
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.4k posts from 5k X accounts over the last 21 days, 9.2k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-04 08:33 UTC. Full method.