Benchmarks / Artificial Analysis

Reported by Artificial Analysis

Artificial Analysis

Median seconds from sending a request to the first token of the answer, including any thinking, measured by Artificial Analysis on the model's usual API.

Last updated 8 Oct 2026

Results dated
8 Oct 2026
Results
133 configurations of 74 models
Unit
seconds
Licence
Artificial Analysis commercial data licence

Time to first answer token: Mistral Large 3

Top 15 of 74 results · seconds, lower is better. Choose a model to highlight it.Clear highlight

  1. 1 Command ACohere 0.40
  2. 2 Gemma 4 E4B (no reasoning)Google 0.41
  3. 3 Claude Haiku 4.5 (no reasoning)Anthropic 0.42
  4. 4 Qwen3.5-9B (no reasoning)Alibaba 0.43
  5. 5 Nemotron 3 Nano 30B A3B (no reasoning)NVIDIA 0.45
  6. 6 Mistral Small 4 (no reasoning)Mistral AI 0.47
  7. 7 Ministral 3 8BMistral AI 0.49
  8. 8 Ministral 3 14BMistral AI 0.51
  9. 9 Ministral 3 3BMistral AI 0.54
  10. 10 Nova Micro 1.0Amazon 0.60
  11. 11 Llama 4 ScoutMeta 0.62
  12. 12 Llama 3.3 70B InstructMeta 0.64
  13. 13 Llama 4 MaverickMeta 0.66
  14. 13 Gemma 4 26B A4B (no reasoning)Google 0.66
  15. 15 Mistral Large 3Mistral AI 0.69

Full results

Artificial Analysis: Time to first answer token, seconds, lower is better
#ModelTime to first answer token
seconds, lower is better		Price
$ per million tokens, in / out
1 Command ACohere
0.40
$2.50 / $10
2 Gemma 4 E4B (no reasoning)Google · best of 2 settings
0.41
–
3 Claude Haiku 4.5 (no reasoning)Anthropic · best of 2 settings
0.42
$1 / $5
4 Qwen3.5-9B (no reasoning)Alibaba · best of 2 settings
0.43
$0.10 / $0.15
5 Nemotron 3 Nano 30B A3B (no reasoning)NVIDIA · best of 2 settings
0.45
$0.05 / $0.20
6 Mistral Small 4 (no reasoning)Mistral AI · best of 2 settings
0.47
$0.15 / $0.60
7 Ministral 3 8BMistral AI
0.49
$0.15 / $0.15
8 Ministral 3 14BMistral AI
0.51
$0.20 / $0.20
9 Ministral 3 3BMistral AI
0.54
$0.10 / $0.10
10 Nova Micro 1.0Amazon
0.60
$0.035 / $0.14
11 Llama 4 ScoutMeta
0.62
$0.18 / $0.59
12 Llama 3.3 70B InstructMeta
0.64
$0.59 / $0.79
13 Llama 4 MaverickMeta
0.66
$0.27 / $0.85
13 Gemma 4 26B A4B (no reasoning)Google
0.66
$0.10 / $0.30
15 Mistral Large 3Mistral AI
0.69
$0.50 / $1.50
16 Qwen3.5-4B (no reasoning)Alibaba · best of 2 settings
0.71
–
17 Claude Sonnet 5.5 (medium reasoning)Anthropic · best of 5 settings
0.76
$2 / $10
17 GPT-5.6 Terra (no reasoning)OpenAI · best of 6 settings
0.76
$2 / $12
19 GPT-6 Luna (no reasoning)OpenAI · best of 5 settings
0.82
$0.10 / $0.50
20 Gemma 4 31B (no reasoning)Google · best of 2 settings
0.83
$0.14 / $0.40
21 DeepSeek-V4.1-Flash (no reasoning)DeepSeek · best of 2 settings
0.85
$0.30 / $1.20
22 Nova 2 Lite (no reasoning)Amazon · best of 4 settings
0.90
$0.30 / $2.50
23 DeepSeek-V4-Pro (0813, no reasoning)DeepSeek · best of 2 settings
0.92
$1.32 / $3.96
24 Qwen3-Omni-30B-A3B-InstructAlibaba
0.94
–
25 Qwen3-Coder-NextAlibaba
0.97
$0.18 / $0.90
26 Qwen3.5-Omni-FlashAlibaba
0.98
–
27 Phi-4Microsoft
1.01
$0.07 / $0.14
28 Qwen3.6-35B-A3B (no reasoning)Alibaba · best of 2 settings
1.04
$0.10 / $1
29 Qwen3.5-122B-A10B (no reasoning)Alibaba · best of 2 settings
1.05
$0.26 / $2.08
30 Qwen3-Next-80B-A3B-InstructAlibaba
1.08
$0.10 / $1.10
31 Qwen3.8-27B (no reasoning)Alibaba · best of 4 settings
1.19
$0.50 / $3
32 Qwen3.5-Omni-PlusAlibaba
1.24
–
33 Gemma 4 12B (no reasoning)Google · best of 2 settings
1.41
–
34 Qwen3.5-397B-A17B (no reasoning)Alibaba · best of 2 settings
1.59
$0.55 / $3.50
35 Claude Opus 5.5 (low reasoning)Anthropic · best of 5 settings
1.69
$4 / $20
35 GPT-6.1 Sol (low reasoning)OpenAI · best of 5 settings
1.69
$2 / $10
37 GPT-6 Astra (low reasoning)OpenAI · best of 5 settings
2.11
$10 / $50
38 Claude Fable 5.1 (low reasoning)Anthropic · best of 5 settings
2.22
$10 / $50
39 Grok 4.7 (low reasoning)xAI · best of 3 settings
3.44
$2 / $6
40 Claude Haiku 5.5 (low reasoning)Anthropic · best of 5 settings
5.08
$0.10 / $0.50
41 o3OpenAI
5.41
$2 / $8
42 Trinity Large ThinkingArcee AI
6.85
$0.25 / $0.80
43 Gemini 3.5 Flash-LiteGoogle
7.29
$0.30 / $2.50
44 gpt-oss-20b (low reasoning)OpenAI · best of 2 settings
8.30
$0.03 / $0.15
45 DeepSeek-V4-Flash-Vision-Exp (max reasoning)DeepSeek
10.1
$0.44 / $1.32
46 gpt-oss-120b (low reasoning)OpenAI · best of 2 settings
11.2
$0.15 / $0.60
47 Qwen3-Next-80B-A3B-ThinkingAlibaba
12.4
$0.15 / $1.20
48 Mistral Medium 3.5Mistral AI
13.1
$1.50 / $7.50
49 Inkling SmallThinking Machines
13.9
$0.45 / $1.20
50 Nemotron 3 Super 120B A12BNVIDIA
14.3
$0.085 / $0.40
51 Gemini 3.8 Flash (high reasoning)Google
14.7
$1.50 / $7.50
52 Inkling (extra-high reasoning)Thinking Machines
15.4
$0.95 / $4.05
53 Gemini 3.1 Pro PreviewGoogle
16.5
$2 / $12
54 GPT-5.5 Instant (2026-06-26)OpenAI
18.1
–
55 Qwen3-Omni-30B-A3B-ThinkingAlibaba
18.9
–
56 Mistral Large 4Mistral AI
19.8
$1.36 / $4.18
57 Reka Flash 3Reka AI
21.9
$0.10 / $0.20
58 Hy3Tencent
24.8
$0.14 / $0.58
59 MiniMax-M3MiniMax
25.2
$0.30 / $1.20
60 Solar Pro 4Upstage
25.5
$0.09 / $0.36
61 Muse Glimmer (high reasoning)Meta
25.6
–
62 Kimi K2.7 CodeMoonshot AI
27.9
$0.95 / $4
63 GLM-5.3 (low reasoning)Z.ai · best of 2 settings
29.0
$1.40 / $4.40
64 GPT-5.3-Codex (extra-high reasoning)OpenAI
36.9
$1.75 / $14
65 Qwen3.7-PlusAlibaba
37.8
$0.32 / $1.28
66 Qwen3.8-Flash-NextAlibaba
38.0
–
67 Granite 4.2 8BIBM
39.1
$0.06 / $0.25
68 MiMo-V2.6-FlashXiaomi
42.2
$0.14 / $0.28
69 GLM-5.3-FlashZ.ai
43.3
$0.15 / $0.50
70 Muse Spark 1.3 (extra-high reasoning)Meta · best of 2 settings
45.4
$1.25 / $4.25
71 Kimi K3 (max reasoning)Moonshot AI · best of 2 settings
47.3
$3 / $15
72 Qwen3.8-2.4T-A95BAlibaba
55.0
$2 / $6
73 MiMo-V2.6-ProXiaomi
55.8
$0.43 / $0.87
73 Qwen3.8-Max (0902)Alibaba
55.8
$2 / $6

Swipe the table sideways for more columns.

Ranks follow the score as shown, so equal numbers share a rank. Each model is shown at its best setting; show every setting. Results as published by Artificial Analysis; we do not re-run them.

What it measures

Median seconds from sending a request to the first token of the answer, including any thinking, measured by Artificial Analysis on the model's usual API.

What it does not measure

Not total time for a long answer, or latency on other hosts.

Source

Artificial Analysis · Methodology

282 results from Artificial Analysis not ranked here · show why

We rank a result only when we can tie it to a specific model you can use. These are left out:

  • Not on sale through the API providers we track: 268
  • A different snapshot or variant from the model we list: 13
  • An unusual combination of settings: 1

Data sourced from Artificial Analysis. Licence: Artificial Analysis commercial data licence.