YOU ARE VIEWING ONE ITEM FROM THE AICRIER FEED

Tencent Hy4 Adds Vision, Computer Use

AICrier tracks AI developer news across Product Hunt, GitHub, Hacker News, YouTube, X, arXiv, and more. This page keeps the article you opened front and center while giving you a path into the live feed.

// WHAT AICRIER DOES

7+

TRACKED FEEDS

24/7

SCRAPED FEED

Short summaries, external links, screenshots, relevance scoring, tags, and featured picks for AI builders.

Tencent Hy4 Adds Vision, Computer Use
OPEN LINK ↗
// 1h agoMODEL RELEASE

Tencent Hy4 Adds Vision, Computer Use

Tencent’s final Hy4 model adds vision and computer-use capabilities to its 770B-parameter, 1M-context system, reaching a reported 96.15% pass rate on WorkBuddy office tasks. It is available through Tencent Cloud and OpenRouter, with weights and a system card coming soon.

// ANALYSIS

Hy4’s significance is its combination of long-context reasoning, multimodality, and GUI/CLI control in an open-model ecosystem. The headline benchmarks are promising but remain Tencent-reported until independent evaluations and weights arrive.

  • –First Hy model with visual understanding and computer use across graphical interfaces and terminals
  • –Targets long-horizon workflows involving documents, spreadsheets, coding, Blender, CAD, and cross-application tasks
  • –Tencent reports an 83.18 WorkBuddy score and first place in 9 of 12 office scenarios
  • –Retains preview pricing while offering up to 1M-token context and 65K-token outputs
  • –Local deployment remains out of reach until Tencent publishes the promised weights and system card
// TAGS
hy4llmopen-weightsmultimodalvisioncomputer-useagent

DISCOVERED

1h ago

2026-11-04

PUBLISHED

1h ago

2026-11-04

RELEVANCE

10/ 10

AUTHOR

thePandaily