Hunyuan HY3
Hunyuan HY3
Switches between fast answers and slow thinking, and charges almost nothing for either.
Updated
The read
HY3 is Tencent's hybrid model: it decides whether a question deserves a quick answer or a reasoning pass, which means you are not paying for deliberation on trivia. Input pricing is the lowest of any model in this index that could reasonably be called frontier-adjacent. It is still labelled a preview, availability outside China is patchy, and the English tooling is sparse.
Where it shines
- Routes between fast and slow thinking automatically
- Lowest input pricing of anything near this quality
- Solid tool calling for an early release
Worth knowing first
- Still a preview, so behaviour is not settled
- Availability and documentation outside China are limited
Index score
80.0
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 26 of 53
- Builds
- LLMs
- Category
- Reasoning
- Output
- Text and images in, text out
- Pricing
- Paid only
- Best for
- Mixed workloads where only some questions deserve thinking
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 02GPT-5.6 SolGPT-5.6 SolProvisionalOpenAI's deepest thinker: the one to hand a problem you cannot solve yourself.Reasoning modelPaid only88.6
- 07DeepSeek V4 ProDeepSeek V4 ProProvisionalA reasoning model at the frontier, published under MIT, priced like an afterthought.Reasoning modelFree tier86.8
- 47Perplexity Sonar ProSonar ProProvisionalA model with the live web wired in, and every claim comes back with a source attached.Reasoning modelPaid only71.6
