Tencent has released the full version of Hy4 on 4 November 2026 — the first model in its Hunyuan series that can understand images and operate a computer through its graphical interface and command line.
What is new
- Vision: Hy4 can read images, charts and documents and reason across multiple images — for the first time in the Hunyuan series
- Computer use: it can operate command-line and graphical interfaces across applications, with targeted training for professional software such as Blender and CAD tools
- Scale: the preview’s model card lists 770 billion total parameters (49 billion active per token) and a 1 million-token context window
- Availability: live now via Tencent Cloud TokenHub, OpenRouter, WorkBuddy and CodeBuddy; API list price is 6 yuan per million input tokens and 18 yuan per million output tokens, with a 50% TokenHub discount until 31 December 2026
Important: weights not out yet
Tencent says it plans to release the model weights and system cards in the coming days — the full model’s weights were not yet published at launch, so it cannot yet be downloaded and run locally. The August preview’s weights were released under the Apache 2.0 licence. Benchmark figures quoted for Hy4 come from Tencent itself, not independent testing.
Related articles on HimachalGovt
- Apple Intelligence Reaches China with Alibaba’s Qwen — https://himachalgovt.com/stories/apple-intelligence-china-alibaba-qwen
- Apple Siri AI Waitlist Ends: Downloads Start Automatically — https://himachalgovt.com/stories/apple-siri-ai-waitlist-ends-downloads-start-automatically
- OpenAI Could Be IPO-Ready in 2027, President Greg Brockman Says — https://himachalgovt.com/stories/openai-ipo-ready-2027-greg-brockman



