Output speed
Tokens generated per second under standard conditions. Higher is faster. Varies significantly by model size, hardware, and provider infrastructure.
Filter by value
Distribution — click row to filter
851→
45.51→
132.31→
67.71→
36.41→
27.71→
108.51→
31.81→
891→
9 entities
| Entity ↕ | Output speed ↕ | Type | Integrations ↓ |
|---|---|---|---|
| Grok 3 | 67.7 | — | 52 |
| o1 | 89 | — | 48 |
| GPT-4 Turbo | 27.7 | — | 47 |
| GPT-4o | 108.5 | — | 37 |
| Mistral Large 2 | 36.4 | — | 30 |
| Claude Sonnet 4.6 | 45.5 | — | 21 |
| GPT-4o mini | 31.8 | — | 17 |
| Claude 3 Haiku | 132.3 | — | 16 |
| Fixture LLM Full | 85 | language-model | 5 |