Hunyuan
53.9
As usual
experience score
▼ 4.4
Confidence: Medium
44 opinions tracked this week · 7-day rolling window
30-day experience score
Version breakdown
Only comments that explicitly name a version contribute to its score.
| Hunyuan Hy3 | mentions only | n=2 | Confidence: Insufficient |
|---|
42 generic family mentions are included in the family score, not any version score.
By experience dimension
| General text | 41 | n=21 | |
|---|---|---|---|
| Coding | not enough data | n=4 | |
| Reasoning | 72 | n=12 | |
| Image / vision | not enough data | n=0 | |
| Roleplay / creative | not enough data | n=0 | |
| Safety & refusals | not enough data | n=0 | |
| Speed & latency | 42 | n=7 | |
| Local deploy | not enough data | n=0 |
Top positive drivers
- strong reasoning Reasoning ×10
- speed quality Speed & latency ×2
- coding performance Coding ×1
- problem solved Coding ×1
- cost quality General text ×1
Top negative drivers
- pricing billing General text ×3
- performance issue Speed & latency ×2
- speed quality Speed & latency ×2
- low signal rant General text ×2
- performance issue General text ×2
What users are saying
🟢 “从推理性能来看，正式版一举超过了同 Size 档位的 DeepSeek V4 Flash 和 MiniMax M3，达到了目前在测的中小尺寸模型最优，极限能力也能越级摸到中大尺寸模型的边”
via zhihu · reasoning · zh
🔴 “v4正式版马上要出了，混元现在出一个不如预览版的模型，谁斩杀谁？”
via zhihu · text_general · zh
🟢 “混元3这个参数能提高到这个水准，我都不敢想ds把后训练搞好了会有多好”
via zhihu · text_general · zh
🔴 “输入命中缓存也贵的多，大概率还是停留在buddy和元宝里用了，api可以选择的更多”
via zhihu · text_general · zh
🟢 “把 Preview 那个被大家调侃的"残疾模型"一路拉到主流水准，进步幅度是能肉眼看出来的”
via zhihu · reasoning · zh