Alternatives?
Run it yourself
estimateBenchmarks
in Chat & KnowledgeFAQ
How much does Hy3 cost?›
$0.13 per 1M input tokens and $0.53 per 1M output tokens; cached input $0.033 per 1M; cheapest via DeepInfra at $0.13 / $0.53. A typical benchmark task costs about $0.072 at default effort.
Is Hy3 free to use?›
There is no free API; the cheapest paid price is $0.13 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Hy3, and what is it best at?›
#71 of 334 in Chat & Knowledge, #63 of 132 in Coding, #63 of 133 in Agents & Tool Use, #52 of 195 in Writing & Language, #52 of 133 in Vision & Documents. Strongest at: Reasoning, Science, Long context.
Is Hy3 better than Gemini 3.5 Flash Lite?›
On Arena Text · non-English, Hy3 scores 1,426.3 vs 1,426.5 — about the same.
How can I use Hy3?›
Through the API of AtlasCloud, DeepInfra, Featherless, GMICloud, Novita and 2 more.
What is the context window of Hy3?›
256K tokens, with up to 128K output tokens.
When was Hy3 released? Is there a newer version?›
Released 2026-07-06. It is the current version in its line.
Is Hy3 open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Hy3 locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Hy3 read images?›
No, it accepts text only.
What is Hy3?›
Hy3 is an open-weights language model from Tencent, released Jul 2026, the successor to Hy3 preview.
What are the best alternatives to Hy3?›
MiMo v2.6 Flash (cheaper), Muse Spark 1.2 (better), GLM 5.3 Flash (open weights).