TTencent · Released Nov 4, 2026
Hy4
The official release of Hy4 delivers comprehensive and substantial improvements over Hy4 Preview, with particularly strong gains in agentic capabilities and performance across a wide range of long-horizon, real-world productivity...
- Top 21%ProofBench
- Released
- Nov 4 2026
- Context window
- 1M 66K out
- Input / 1M tokens
- $0.42
- Output / 1M tokens
- $1.25
Where to run it
8 providers compared
Included in Tencent Token Plan, OpenCode Go. Cheapest: Tencent, Vercel AI Gateway at $0.42 / $1.25 per 1M input / output tokens — 50% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| Tencent Token Plansubscription plan | Plan | Plan | Plan | 1M | 66K | — | Plan |
| OpenCode Gosubscription plan | Plan | Plan | Plan | 1M | 66K | — | Plan |
| TencentFP4 | $0.42 | $1.25 | $0.021 | 1M | 66K | 100%24h | Cheapest |
| Vercel AI Gateway | $0.42 | $1.25 | $0.021 | 1.02M | 64K | — | Cheapest |
| NanoGPT | $0.46 | $1.39 | $0.023 | 1M | 66K | — | +11% |
| Eden AI | $0.83 | $2.50 | $0.042 | 1M | 66K | — | +100% |
| Kilo Gateway | $0.83 | $2.50 | $0.042 | 1M | 66K | — | +100% |
| Tencent TokenHub | $0.83 | $2.50 | $0.042 | 1M | 66K | — | +100% |
Value
Price vs. performance
Hy4 Other Tencent Other labs Value frontier · ProofBench vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
1 benchmark score
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| ProofBenchMaths · Graduate maths proofs verified in Lean 4 | 75.0% | #15of 72 | Claude Sonnet 5.5100.0% |
Specs
Details
- Model ID
tencent/hy4- Developer
- Tencent
- Release date
- Nov 4, 2026 ·
- Context window
- 1,048,576 tokens
- Max output
- 65,536 tokens
- API price
- $0.42 / $1.25 per 1M input / output tokens
- Providers
- 8
- Input types
- text, image
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; 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.