aifollow.news
Back IT之家 科技新闻
IT之家 科技新闻· · Original publication time AI score47

Tencent Hunyuan expands Hy4 capacity for WorkBuddy and extends Hy3 free accessMachine translation

腾讯混元扩容 WorkBuddy 的 Hy4 推理集群,延长 Hy3 限免

AI-assisted summary

Tencent Hunyuan and WorkBuddy said many users hit task limits on Hy4’s first day in WorkBuddy. They have increased Hy4 inference cluster capacity tenfold and raised the free trial allowance, though limits may still apply at peak times. Free access to Hy3 in WorkBuddy now runs until November 15, 2026, at 23:59.

Article · AI translation

Thanks to IT Home reader 不一样的体验 for the tip!

IT Home reported on November 5 that Tencent Hunyuan and WorkBuddy posted today that, since Hunyuan Hy4 was first made available on WorkBuddy on November 4, 2026, it has received enthusiastic responses from users and developers who tried it. On its first day, many users reached their limits in the WorkBuddy task queue.

The company said it had urgently expanded the Hy4 inference cluster tenfold and increased users’ free trial limits to safeguard the user experience, and would continue to adjust resources dynamically. However, given the limited total supply of high-end computing capacity and concentrated demand at peak times, limits may still be reached during certain periods. If that happens, it recommends the following temporary options:

Switch to Hy3 or another model to continue using the service: the current free limited-time access to Hy3 on WorkBuddy will be extended until 23:59 on November 15, 2026.

For non-urgent needs, use the service during off-peak evening hours.

As IT Home previously reported, Tencent Hunyuan released the official version of its latest-generation large language model, Hy4, yesterday. Compared with Hy4 preview, Hy4 substantially expands the scale of pretraining and post-training, introduces visual capabilities for the first time, and significantly improves its overall abilities in instruction following, long-text comprehension, long-horizon tasks, deep research and other areas.

The company said that, with another comprehensive leap in model capabilities, Hy has joined the ranks of the strongest open-source models for the first time. Hy4 will open-source its model weights and publish a model system card soon, while actively pursuing more responsible open source, with a focus on risks including cybersecurity and biosecurity.

Found an error? Send a correction