Tencent’s flagship model learns to see and use a computer
Hy4 is Tencent’s latest flagship model, aimed at office tasks, financial analysis and game development. New since the preview, per Tencent: visual understanding for charts and images, and computer use across command line and graphical interfaces, with training on Blender and CAD. Tencent says the internal hallucination rate fell from 12.95% to 7.2%. Available on Tencent Cloud TokenHub, OpenRouter and Tencent apps. Weights are promised in the coming days.
Hy4 is Tencent’s flagship model, and this full release follows the Hy4 Preview from August. It adds two big things: it can see, and it can use a computer.
Here’s what’s new:
Visual understanding for fine-grained images, complex charts and math diagrams
Computer use across command-line and graphical interfaces, with training on professional software like Blender and CAD
Better instruction following, long-context comprehension and long, multi-step agent tasks
An internal hallucination rate that fell from 12.95% to 7.2%, and errors from unsupported assumptions that fell from 23.1% to 5.6%
On Tencent’s own tests, Hy4 scored 3.04 in a blind financial evaluation (213 professionals, 579 tasks), ahead of DeepSeek-V4.1-Flash-1017 at 2.97 and Kimi K3.1 Pro and GLM-5.3 at 2.68. It also scored 83.18 on WorkBuddy with a 96.15% pass rate. These are Tencent’s own results, not independent tests.
It’s aimed at office tasks, financial analysis and game development, and Tencent says product teams helped train and evaluate it against real workflows.
Availability: it’s on Tencent Cloud TokenHub, OpenRouter, Hermes and OpenCode, and inside Tencent’s apps WorkBuddy, CodeBuddy, Yuanbao, ima and WeCom’s AI assistant. WorkBuddy and CodeBuddy are free from November 4 to 15, and TokenHub has 50% off through December 31. The model weights aren’t out yet, and Tencent says they’ll come in the coming days, along with system cards.
Good fit if you build agents or workflows that need a model that can read charts and documents and operate software, and you want to try it before the weights land.
About Tencent Hy4 on Product Hunt
“Tencent’s flagship model learns to see and use a computer”
Tencent Hy4 launched on Product Hunt on November 5th, 2026 and earned 82 upvotes and 1 comments, placing #15 on the daily leaderboard. Hy4 is Tencent’s latest flagship model, aimed at office tasks, financial analysis and game development. New since the preview, per Tencent: visual understanding for charts and images, and computer use across command line and graphical interfaces, with training on Blender and CAD. Tencent says the internal hallucination rate fell from 12.95% to 7.2%. Available on Tencent Cloud TokenHub, OpenRouter and Tencent apps. Weights are promised in the coming days.
On the analytics side, Tencent Hy4 competes within Artificial Intelligence — topics that collectively have 483k followers on Product Hunt. The dashboard above tracks how Tencent Hy4 performed against the three products that launched closest to it on the same day.
Who hunted Tencent Hy4?
Tencent Hy4 was hunted by Kumar Abhishek. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
For a complete overview of Tencent Hy4 including community comment highlights and product details, visit the product overview.
Kumar Abhishek
Hy4 is Tencent’s flagship model, and this full release follows the Hy4 Preview from August. It adds two big things: it can see, and it can use a computer.
Here’s what’s new:
Visual understanding for fine-grained images, complex charts and math diagrams
Computer use across command-line and graphical interfaces, with training on professional software like Blender and CAD
Better instruction following, long-context comprehension and long, multi-step agent tasks
An internal hallucination rate that fell from 12.95% to 7.2%, and errors from unsupported assumptions that fell from 23.1% to 5.6%
On Tencent’s own tests, Hy4 scored 3.04 in a blind financial evaluation (213 professionals, 579 tasks), ahead of DeepSeek-V4.1-Flash-1017 at 2.97 and Kimi K3.1 Pro and GLM-5.3 at 2.68. It also scored 83.18 on WorkBuddy with a 96.15% pass rate. These are Tencent’s own results, not independent tests.
It’s aimed at office tasks, financial analysis and game development, and Tencent says product teams helped train and evaluate it against real workflows.
Availability: it’s on Tencent Cloud TokenHub, OpenRouter, Hermes and OpenCode, and inside Tencent’s apps WorkBuddy, CodeBuddy, Yuanbao, ima and WeCom’s AI assistant. WorkBuddy and CodeBuddy are free from November 4 to 15, and TokenHub has 50% off through December 31. The model weights aren’t out yet, and Tencent says they’ll come in the coming days, along with system cards.
Good fit if you build agents or workflows that need a model that can read charts and documents and operate software, and you want to try it before the weights land.