Benchmarks · speed
Provider throughput
Median provider output throughput in tokens per second across OpenRouter endpoints (3d).
180 models · 180 results · data as of 2026-10-29 · openrouter.ai ↗
Leaderboard
Muse Spark 1.2 max
All results
180 rows| # | Model | Effort | Provider throughput |
|---|---|---|---|
| 1 | Muse Spark 1.2 | max | 239.5 |
| 2 | Muse Spark 1.1 | max | 202.0 |
| 3 | Gemini 2.5 Flash Lite | none | 158.5 |
| 4 | Muse Spark 1.3 | max | 158.5 |
| 5 | Haiku 5.5 | max | 155.0 |
| 6 | Grok Build 0.1 | 143.8 | |
| 7 | gpt-oss-120b | high | 140.5 |
| 8 | Gemini 3.1 Flash-Lite | high | 135.9 |
| 9 | Gemini 3.5 Flash-Lite | xhigh | 128.5 |
| 10 | Llama 3.2 3b | none | 126.5 |
| 11 | GPT-5.6 Luna Pro | 126.0 | |
| 12 | DeepSeek V4.1 Flash 1017 | max | 125.3 |
| 13 | Inkling Small | max | 121.0 |
| 14 | Gemini 3.5 Flash | xhigh | 120.0 |
| 15 | Gemini 3.6 Flash | xhigh | 117.0 |
| 16 | GPT-4.1 Nano | high | 115.0 |
| 17 | Mistral Medium 3 | none | 113.9 |
| 18 | Llama 3.1 8b | medium | 112.5 |
| 19 | Mistral Small 4 | high | 110.5 |
| 20 | Gemini 2.5 Flash | medium | 105.5 |
| 21 | DeepSeek V4 Flash | max | 105.0 |
| 22 | DeepSeek V4.1 Flash | max | 105.0 |
| 23 | GPT-5.6 Sol Pro | max | 104.8 |
| 24 | Llama 3.2 1b | none | 103.8 |
| 25 | Qwen 2.5 7B | 102.5 | |
| 26 | GPT-6 Luna Pro | max | 101.0 |
| 27 | Mistral Medium 3.5 | high | 100.0 |
| 28 | GPT-5.4 Mini | xhigh | 99.5 |
| 29 | GPT-5.6 Terra Pro | 97.5 | |
| 30 | Mistral Medium 3.1 | none | 96.5 |