Benchmarks · speed
Output Speed (tokens/s)
Median output tokens per second measured by AA (median across providers, or first-party API).
117 models · 238 results · data as of 2026-10-29 · artificialanalysis.ai ↗
Leaderboard
Gemini 3.5 Flash-Lite high
All results
238 rows| # | Model | Effort | Output Speed (tokens/s) |
|---|---|---|---|
| 1 | Gemini 3.5 Flash-Lite | high | 454.7 |
| 2 | Muse Spark 1.2 | xhigh | 332.3 |
| 3 | Gemini 2.5 Flash Lite | none | 309.8 |
| 4 | Muse Spark 1.1 | xhigh | 307.5 |
| 5 | Muse Spark 1.3 | xhigh | 295.1 |
| 6 | Nemotron 3.5 Lightning | 294.5 | |
| 7 | GPT-4.1 Nano | none | 290.5 |
| 8 | DeepSeek V4.1 Flash | none | 262.2 |
| 9 | gpt-oss-120b | low | 259.8 |
| 10 | o3 mini | high | 258.1 |
| 11 | Ling 3.0 Flash | 243.0 | |
| 12 | gpt-oss-20b | low | 240.9 |
| 13 | Haiku 5.5 | max | 240.0 |
| 14 | o3 mini | 233.8 | |
| 15 | GPT-5.4 Mini | xhigh | 222.7 |
| 16 | Ling 3.1 Flash | 222.6 | |
| 17 | Gemini 3.5 Flash | medium | 214.8 |
| 18 | Gemini 3.5 Flash | high | 211.4 |
| 19 | Gemini 3.6 Flash | high | 211.0 |
| 20 | Muse Spark 1.3 | max | 208.5 |
| 21 | Inkling | xhigh | 207.2 |
| 22 | Gemini 2.5 Flash | none | 206.4 |
| 23 | Inkling Small | 204.0 | |
| 24 | Qwen 3.7 Max | max | 203.6 |
| 25 | GPT-5.4 Mini | medium | 200.6 |
| 26 | gpt-oss-120b | high | 200.4 |
| 27 | GPT-5 Codex | high | 199.3 |
| 28 | Gemini 3.5 Flash | minimal | 198.3 |
| 29 | Haiku 5.5 | xhigh | 194.7 |
| 30 | Haiku 5.5 | high | 191.7 |