TTencent · Released Jul 6, 2026
Hy3
Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...
- Released
- Jul 6 2026
- Context window
- 262K 128K out
- Input / 1M tokens
- $0.13
- Output / 1M tokens
- $0.53
Where to run it
5 providers compared
Cheapest: Tencent at $0.13 / $0.53 per 1M input / output tokens — 34% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| TencentFP8 | $0.13 | $0.53 | $0.033 | 262K | 128K | 100%24h | Cheapest |
| Novita | $0.14 | $0.58 | $0.035 | 262K | 236K | 100%24h | +8% |
| GMICloudBF16 | $0.14 | $0.58 | $0.035 | 262K | 236K | 100%24h | +8% |
| Phala | $0.15 | $0.64 | $0.04 | 262K | 236K | 99.8%24h | +18% |
| AtlasCloudFP8 | $0.20 | $0.80 | $0.05 | 262K | 131K | 100%24h | +52% |
Benchmarks
Benchmark scores
No benchmark scores yet — Epoch AI hasn't evaluated Hy3. Scores appear here automatically once it does.
Specs
Details
- Model ID
tencent/hy3- Developer
- Tencent
- Release date
- Jul 6, 2026 ·
- Context window
- 262,144 tokens
- Max output
- 128,000 tokens
- API price
- $0.13 / $0.53 per 1M input / output tokens
- Providers
- 5
- Input types
- text
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Free tier
- No
Release date, prices and providers from OpenRouter's catalog; buying directly from a provider can cost a little differently. Uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.