Fireworks vs Novita
15 vs 67 models, 13 shared
Shared models
| Model | Fireworks $/1M out | Fireworks tok/s | Novita $/1M out | Novita tok/s |
|---|---|---|---|---|
| MiniMax-M2.1 | 84 tok/s | 27 tok/s | ||
| MiniMax-M2.5 | 29 tok/s | 16 tok/s | ||
| Qwen3-VL-30B-A3B-Instruct | 162 tok/s | 118 tok/s | ||
| Qwen3-VL-30B-A3B-Thinking | 132 tok/s | 87 tok/s | ||
| DeepSeek-V3.1 | 79 tok/s | 36 tok/s | ||
| DeepSeek-V3.2 | 81 tok/s | 29 tok/s | ||
| Llama-3.3-70B-Instruct | 109 tok/s | 27 tok/s | ||
| Kimi-K2-Instruct-0905 | 40 tok/s | 24 tok/s | ||
| Kimi-K2-Thinking | 49 tok/s | 28 tok/s | ||
| Kimi-K2.5 | 67 tok/s | 102 tok/s | ||
| gpt-oss-120b | 90 tok/s | 51 tok/s | ||
| gpt-oss-20b | 143 tok/s | 97 tok/s | ||
| GLM-5 | 70 tok/s | 39 tok/s |