Meta Llama 3.3 70B Instruct vs Tencent Hy-MT2-7B
Side by side from live data: prices, limits, who serves each model and how reliably, and how they do on our chess arena.
In short
- Hy-MT2-7B is 1.2× cheaper at a typical mix of 3 input tokens per output token.
- Llama 3.3 70B Instruct accepts longer prompts: 131K vs 8K tokens.
- Hy-MT2-7B is the newer model (released 2026-08-19).
| Llama 3.3 70B Instruct | Hy-MT2-7B | |
|---|---|---|
| Vendor | Meta | Tencent |
| Released | 2024-12-06 | 2026-08-19 |
| Input / 1M | $0.10 | $0.074 |
| Output / 1M | $0.32 | $0.295 |
| Cached input / 1M | — | — |
| Blended / 1M (3:1) | $0.155 | $0.129 |
| Context | 131K | 8K |
| Max output | 16K | 4K |
| Accepts | text | text |
| Reasoning | No | No |
| Providers on OpenRouter | 10 | 1 |
| Cheapest provider, output | $0.32 | $0.295 |
| Best provider uptime, 24 h | 100% | 99.68% |
| Chess arena rating | not playing | not playing |
| Vendor API status | — | — |
Cost calculator
Llama 3.3 70B Instruct
$10.80 per month$0.36 per day
Hy-MT2-7B
$8.87 per month$0.30 per day
Hy-MT2-7B saves $1.94 a month on this workload.
At list prices per 1M tokens; prompt caching and batch discounts are not included.
Compare Llama 3.3 70B Instruct with