Hy3-preview (Non-reasoning)
AI model by Tencent. Real-time pricing and benchmark data.
Benchmarks
Editorial Analysis
Profile. Hy3-preview (Non-reasoning) is a budget-tier option from Tencent at $0.21/M output tokens. Despite the low price, it scores competitively on measurable benchmarks — making it a candidate for high-volume production workloads where unit cost dominates.
Cost model. For a workload of 1M input tokens + 100K output tokens, Hy3-preview (Non-reasoning) costs $0.08 per million input-equivalent requests. Multiply by your monthly request volume to project the bill, and compare against the speed/correctness profile above to decide if the trade-off is worth it for your workload.
Editorial summary generated from public pricing and benchmark data (Artificial Analysis API). All numbers are sourced — see the Artificial Analysis methodology for benchmark definitions. Last refreshed 2026-08-06.
Compare with similar models
| Model | Input | Output | Speed |
|---|---|---|---|
Hy3-preview (Non-reasoning)Current | $0.06 | $0.21 | — |
| NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning) | $0.05 | $0.20 | 197 tok/s |
| Qwen3.5 9B (Reasoning) | $0.14 | $0.20 | 71 tok/s |
| NVIDIA Nemotron Nano 9B V2 (Non-reasoning) | $0.05 | $0.20 | 170 tok/s |
| Qwen2.5 Turbo | $0.05 | $0.20 | 107 tok/s |
| gpt-oss-20b (high) | $0.07 | $0.20 | 165 tok/s |