As of today, it has fallen to 8/9th on the rankings. I don't see a reason where you would use this model over competitors. However, price economics are bit confusing, as currently the effective input price of Hy3 via OpenRouter is now the same as DeepSeek-hosted DeepSeek Flash V4.
I had to stop using it because I was getting rate limited like crazy. Probably why it has dropped.
Catloafdev 39 minutes ago [-]
Curious how people feel about this compared to DS4 Flash, given they are pretty close in size. Also curious how well it holds up to heavy quantization.
DS4 Flash can currently run reasonably well on systems with ~96gb+ RAM, I wonder if Hy3 can compete there.
UncleOxidant 26 minutes ago [-]
That's a 2-bit quant of DS4 flash. You're probably better off running Qwen3.6-27B at Q8.
Catloafdev 55 seconds ago [-]
For most coding or agentic tasks, Qwen 3.6 27B likely outperforms, yes.
For 'general intelligence', DS4 Flash seems to be a noticeable step up still.
spmurrayzzz 16 minutes ago [-]
I think its good advice to test both on your own evals for sure, but the MoE parameters are already natively FP4 in ds4. Dropping to 2bpw isn't as big of a loss as it seems (and as corroborated by antirez's work).
Its also only 13B active, so your decode speed would be nearly 2x that of Qwen3.6-27B. So there are other latent benefits as well.
sosodev 21 minutes ago [-]
I suspect it would depend on the task. DS4-flash does, as previously mentioned, handle quantization very well. Even at 2-bit it's still very coherent.
wolttam 29 minutes ago [-]
Hy3 lacks the DSv4 architecture's KV Cache efficiency.
Whereas I can run DSv4 Flash on a pair of DGX Sparks and have enough memory left over for 3M tokens of KV cache, with Hy3 (quantized to FP4), there is only room for ~130K tokens of KV cache.
ignoramous 14 minutes ago [-]
Lower context window notwithstanding, Hy3's coding benchmarks hold their own against DeepSeek v4 Pro & MiMo v2.5 Pro. That's quite something for a model priced like DeepSeek v4 Flash & MiMo v2.5 (for non-cached tokens), which are 3x cheaper than their respective Pro variants.
wolttam 7 minutes ago [-]
It's impressive indeed. I would also expect the next checkpoint of DSv4 Flash to come in somewhere at this level (DeepSeek has had over 2 months to continue training since it released).
It's exciting that the open models continue to get better and more efficient across the board!
nunodonato 38 minutes ago [-]
DS4-Flash is not only "significantly" smaller, it will also benefit from a lot more speed thanks to DSpark
Oh, it is. I was looking at the Huggingface repo which listed the lower number at the top of the page, looks like that's wrong.
minraws 48 minutes ago [-]
I tried out the model it's pretty great, better than ~~gpt5.4~~ gpt-5.4-mini perhaps, atleast close enough to sonnet 5 in performance that I didn't notice much of a gap.
Not really at gpt 5.5 tier though, and probably below glm 5.2...
But most of all it just works for me for most things I tried and it's exceedingly cheap so there is no reason not to use it, if you need a foss model.
Edited: gpt-5.4-mini not the base gpt-5.4
theplumber 19 minutes ago [-]
I think you’ve got the models wrong…gpt-5.4? I doubt there is any open source mode matching it. Maybe in a year
mgrandl 10 minutes ago [-]
GLM 5.2 already matches GPT-5.4 easily.
minraws 12 minutes ago [-]
Yeah I meant gpt-5.4-mini, but GLM 5.2 is pretty close to gpt-5.4 base, and much better than it when it comes to design stuff.
cbg0 19 minutes ago [-]
Hy3 DeepSWE - 28%
GPT5.4 xhigh DeepSWE - 52%
A lot of contaminated benchmarks in the blog post about Hy3, needs real testing though I have a distinct feeling it's benchmaxxed like a lot of Chinese models.
doawoo 7 minutes ago [-]
That UI demo page is… really quite janky.
nshotton 45 minutes ago [-]
This model is shockingly small for how capable it is. its a little bit bigger than deepseekV4 flash but around as capable if not more on some benchmarks than V4 pro, i wouldnt be surprised if this becomes a popular local model.
andai 39 minutes ago [-]
I've been wondering about that. GLM-5.2 is also half the size of DeepSeek V4 Pro. (But costs roughly twice as much.)
I looked into DeepSeek's architecture a little bit and the main focus was how can we save as much money as possible. They did a lot of cost cutting with the attention mechanisms. This allowed them to offer an insanely cheap price even on massive contexts, but seems to have come at the cost of performance?
At least, that's my guess, when I see smaller models costing more and outperforming, I think, "they must have denser attention?"
nunodonato 44 minutes ago [-]
hardly, its still quite big unless by "local" you mean people that spend many thousands on rigs :)
nshotton 41 minutes ago [-]
Yeah i shouldve been more clear, a model of this size could run on 2 dgx sparks so out of the range of a lot of the typical consumer sure, but I think there is definitely a market for that size
IshKebab 38 minutes ago [-]
> Hy3 has 295B parameters in total. To serve it on 8 GPUs, we recommend using H20-3e or other GPUs with larger memory capacity.
I would.
throwaway2027 33 minutes ago [-]
Quite interesting to see them and Meta and others release before OpenAI supposedly is to release GPT 5.6 today, would it be better to release it before or after? Calm before the storm type of thing?
Kye 4 minutes ago [-]
Release before and GPT 5.6 has to be better enough than early experiences with other new models to warrant the premium.
james2doyle 35 minutes ago [-]
Been using this and GLM 5.2 back and forth. I like the speed of Hy3. Also seems very happy to follow instructions. Still haven’t found any open models that follow instructions as good as Mimo v2 pro though
handzhiev 44 minutes ago [-]
It's a very good model for this size and price. I tried it with a couple of small tasks - just an year ago this would be the level of the leading models.
smir 52 minutes ago [-]
Visited the link thinking it's for hy lang, found it's another llm from tencent, anyway it's nice read
nunodonato 45 minutes ago [-]
Very impressive model for its size
Rendered at 16:51:40 GMT+0000 (Coordinated Universal Time) with Vercel.
As of today, it has fallen to 8/9th on the rankings. I don't see a reason where you would use this model over competitors. However, price economics are bit confusing, as currently the effective input price of Hy3 via OpenRouter is now the same as DeepSeek-hosted DeepSeek Flash V4.
https://openrouter.ai/tencent/hy3-preview
https://openrouter.ai/deepseek/deepseek-v4-flash
DS4 Flash can currently run reasonably well on systems with ~96gb+ RAM, I wonder if Hy3 can compete there.
For 'general intelligence', DS4 Flash seems to be a noticeable step up still.
Its also only 13B active, so your decode speed would be nearly 2x that of Qwen3.6-27B. So there are other latent benefits as well.
Whereas I can run DSv4 Flash on a pair of DGX Sparks and have enough memory left over for 3M tokens of KV cache, with Hy3 (quantized to FP4), there is only room for ~130K tokens of KV cache.
It's exciting that the open models continue to get better and more efficient across the board!
Edit: fixed, got bad info
Not really at gpt 5.5 tier though, and probably below glm 5.2...
But most of all it just works for me for most things I tried and it's exceedingly cheap so there is no reason not to use it, if you need a foss model.
Edited: gpt-5.4-mini not the base gpt-5.4
GPT5.4 xhigh DeepSWE - 52%
A lot of contaminated benchmarks in the blog post about Hy3, needs real testing though I have a distinct feeling it's benchmaxxed like a lot of Chinese models.
I looked into DeepSeek's architecture a little bit and the main focus was how can we save as much money as possible. They did a lot of cost cutting with the attention mechanisms. This allowed them to offer an insanely cheap price even on massive contexts, but seems to have come at the cost of performance?
At least, that's my guess, when I see smaller models costing more and outperforming, I think, "they must have denser attention?"
I would.