Hands-on: Tencent's Hy4 matches or beats rivals across five agentic coding tests
CodeByPoonam · x · 2026-11-10
Blogger CodeByPoonam benchmarked Tencent's Hy4 against rival tools with identical prompts and settings, and summed up the results:
- Decks: a 20+ page customer-service training deck from one prompt (product tables, roleplays, compliance scripts) — Hy4 matched or beat others on layout
- Computer use: recreated a logo in Excalidraw using only mouse actions in Chrome, reading the screen, picking tools and self-correcting
- 3D & web builds: an interactive Three.js mechanical watch movement you can disassemble part by part, a 3D paper-cut "Monkey King" storybook with parallax, and a single-prompt pixel art editor with layers, onion skinning and GIF export
Verdict: Hy4 produced more complete results on first attempts and performed strongly across all five tests.
More from coding & agent
- Leetcode interviews will die as agentic coding rounds become the new hiring standard — rachittshah · 2026-11-10
- 10 open-source GitHub repos for voice AI: from local ASR to zero-shot voice cloning — Shruti_0810 · 2026-11-10
- As Arc sunsets, a dev builds Earful, an AI tool that aggregates product reviews into verdicts — BLCNYY · 2026-11-10
- Claude's Chrome Extension Slammed for Being Unable to Control Already-Open Tabs — Angaisb_ · 2026-11-10
- Nebius releases Longtail-SWE: 202 repo-level SWE tasks across 13 underrepresented languages — nebius · 2026-11-10
- Self-hosting OpenMausBot for half a day: it runs, but don't trust it with real work — moutsea · 2026-11-10