Benchmarks · speed
Latency (Time to First Token)
Median seconds from request to first streamed chunk as measured by AA; may be a reasoning token. This is not time to first answer.
117 models · 238 results · data as of 2026-10-29 · artificialanalysis.ai ↗
Leaderboard
Gemini 2.5 Flash Lite none
All results
238 rows| # | Model | Effort | Score |
|---|---|---|---|
| 1 | Gemini 2.5 Flash Lite | none | 0.3s |
| 2 | Gemini 2.5 Flash | none | 0.4s |
| 3 | Nemotron 3.5 Lightning | 0.6s | |
| 4 | GPT-5.4 Nano | none | 0.7s |
| 5 | Grok 4.3 | none | 0.7s |
| 6 | gpt-oss-20b | high | 0.7s |
| 7 | gpt-oss-120b | low | 0.8s |
| 8 | Sonnet 5 | none | 0.8s |
| 9 | GPT-5.4 Mini | none | 0.8s |
| 10 | GPT-5.6 Luna | none | 0.8s |
| 11 | GPT-5.6 Terra | none | 0.8s |
| 12 | GPT-5.2 | none | 0.8s |
| 13 | GPT-5 Nano | minimal | 0.8s |
| 14 | Llama 4 Scout | none | 0.8s |
| 15 | Gemini 3.5 Flash | minimal | 0.8s |
| 16 | gpt-oss-120b | high | 0.8s |
| 17 | Qwen 3.5 9B | none | 0.8s |
| 18 | GPT-6 Luna | none | 0.9s |
| 19 | gpt-oss-20b | low | 0.9s |
| 20 | GPT-4.1 Mini | none | 0.9s |
| 21 | DeepSeek V4.1 Flash | max | 0.9s |
| 22 | Sonnet 4.6 | none | 0.9s |
| 23 | Mistral Small 3 | none | 0.9s |
| 24 | Mistral Small 4 | none | 0.9s |
| 25 | Qwen 3.5 4B | 0.9s | |
| 26 | Mistral Small 3.2 | none | 0.9s |
| 27 | Mistral Small 4 | 0.9s | |
| 28 | DeepSeek V4.1 Flash | none | 0.9s |
| 29 | Llama 4 Maverick | none | 1.0s |
| 30 | GPT-6 Sol | none | 1.0s |