Benchmarks / Artificial Analysis

Reported by Artificial Analysis

Artificial Analysis

Median tokens a second while the model writes its answer, measured by Artificial Analysis on the model's usual API over the past weeks.

Last updated 20 Oct 2026

Results dated
20 Oct 2026
Results
137 configurations of 78 models
Unit
tokens per second
Licence
Artificial Analysis commercial data licence

Output speed: Muse Spark Open

Top 15 of 137 results · tokens per second, higher is better. Choose a model to highlight it.Clear highlight

  1. 1 Nemotron 3.5 Super VLNVIDIA 411
  2. 2 Gemini 3.5 Flash-LiteGoogle 398
  3. 3 DeepSeek-V4.1-Flash (no reasoning)DeepSeek 369
  4. 4 Trinity Large ThinkingArcee AI 362
  5. 5 Beam 1.0 501B A23BReflection 323
  6. 6 gpt-oss-20b (low reasoning)OpenAI 277
  7. 7 gpt-oss-20b (high reasoning)OpenAI 263
  8. 8 Nova Micro 1.0Amazon 260
  9. 9 gpt-oss-120b (high reasoning)OpenAI 259
  10. 10 gpt-oss-120b (low reasoning)OpenAI 258
  11. 11 Muse Spark 1.3 (extra-high reasoning)Meta 254
  12. 12 Claude Haiku 5.5 (max reasoning)Anthropic 240
  13. 13 Nemotron 3 Nano 30B A3B (reasoning on)NVIDIA 238
  14. 14 Nova 2 Lite (no reasoning)Amazon 233
  15. 15 Inkling (extra-high reasoning)Thinking Machines 232
  16. 42 Muse Spark Open (extra-high reasoning)Meta 142

Full results

Artificial Analysis: Output speed, tokens per second, higher is better
#ModelOutput speed
tokens per second, higher is better		Price
$ per million tokens, in / out
1 Nemotron 3.5 Super VLNVIDIA
411
–
2 Gemini 3.5 Flash-LiteGoogle
398
$0.30 / $2.50
3 DeepSeek-V4.1-Flash (no reasoning)DeepSeek
369
$0.30 / $1.20
4 Trinity Large ThinkingArcee AI
362
$0.25 / $0.80
5 Beam 1.0 501B A23BReflection
323
–
6 gpt-oss-20b (low reasoning)OpenAI
277
$0.03 / $0.15
7 gpt-oss-20b (high reasoning)OpenAI
263
$0.03 / $0.15
8 Nova Micro 1.0Amazon
260
$0.035 / $0.14
9 gpt-oss-120b (high reasoning)OpenAI
259
$0.15 / $0.60
10 gpt-oss-120b (low reasoning)OpenAI
258
$0.15 / $0.60
11 Muse Spark 1.3 (extra-high reasoning)Meta
254
$1.25 / $4.25
12 Claude Haiku 5.5 (max reasoning)Anthropic
240
$0.10 / $0.50
13 Nemotron 3 Nano 30B A3B (reasoning on)NVIDIA
238
$0.05 / $0.20
14 Nova 2 Lite (no reasoning)Amazon
233
$0.30 / $2.50
15 Inkling (extra-high reasoning)Thinking Machines
232
$0.95 / $4.05
15 Nemotron 3 Nano 30B A3B (no reasoning)NVIDIA
232
$0.05 / $0.20
17 DeepSeek-V4.1-Flash (max reasoning)DeepSeek
231
$0.30 / $1.20
17 Qwen3.5-Omni-FlashAlibaba
231
–
19 Nova 2 Lite (high reasoning)Amazon
220
$0.30 / $2.50
20 DeepSeek-V4.1-Flash-1017 (max reasoning)DeepSeek
212
$0.30 / $1.20
21 Ministral 3 3BMistral AI
209
$0.10 / $0.10
22 Nova 2 Lite (medium reasoning)Amazon
200
$0.30 / $2.50
23 Claude Haiku 5.5 (extra-high reasoning)Anthropic
192
$0.10 / $0.50
24 DeepSeek-V4-Flash-Vision-Exp (max reasoning)DeepSeek
191
$0.44 / $1.32
25 Inkling SmallThinking Machines
181
$0.45 / $1.20
26 Nova 2 Lite (low reasoning)Amazon
177
$0.30 / $2.50
27 Mistral Small 4Mistral AI
176
$0.15 / $0.60
28 Claude Haiku 5.5 (high reasoning)Anthropic
169
$0.10 / $0.50
29 Mistral Small 4 (no reasoning)Mistral AI
166
$0.15 / $0.60
30 Qwen3-Next-80B-A3B-ThinkingAlibaba
163
$0.15 / $1.20
31 MiniMax M3.1 FlashMiniMax
159
$0.15 / $0.60
32 Mistral Medium 3.5Mistral AI
157
$1.50 / $7.50
32 Claude Haiku 5.5 (low reasoning)Anthropic
157
$0.10 / $0.50
34 Claude Haiku 5.5 (medium reasoning)Anthropic
155
$0.10 / $0.50
34 o3OpenAI
155
$2 / $8
36 Muse Wonder (max reasoning)Meta
153
$4 / $20
37 GPT-5.5 Instant (2026-06-26)OpenAI
152
–
38 Gemma 4 12BGoogle
148
–
39 MiniMax-M3MiniMax
146
$0.30 / $1.20
40 Gemma 4 12B (no reasoning)Google
145
–
41 Qwen3.5-122B-A10B (no reasoning)Alibaba
144
$0.26 / $2.08
42 Muse Glimmer (high reasoning)Meta
142
–
42 Muse Spark Open (extra-high reasoning)Meta
142
–
44 Gemini 3.8 Flash (high reasoning)Google
141
$1.50 / $7.50
45 GPT-5.6 Terra (max reasoning)OpenAI
140
$2 / $12
46 Claude Sonnet 5.5 (max reasoning)Anthropic
137
$2 / $10
47 Qwen3.6-35B-A3B (no reasoning)Alibaba
135
$0.10 / $1
48 Qwen3.6-35B-A3BAlibaba
134
$0.10 / $1
49 Llama 4 ScoutMeta
132
$0.18 / $0.59
49 Qwen3.5-122B-A10BAlibaba
132
$0.26 / $2.08
51 Gemini 3.1 Pro PreviewGoogle
130
$2 / $12
52 GPT-6 Luna (max reasoning)OpenAI
127
$0.10 / $0.50
53 GPT-6 Luna (high reasoning)OpenAI
122
$0.10 / $0.50
54 Qwen3-Next-80B-A3B-InstructAlibaba
121
$0.10 / $1.10
55 Muse Spark 1.3 (max reasoning)Meta
119
$1.25 / $4.25
55 GPT-5.3-Codex (extra-high reasoning)OpenAI
119
$1.75 / $14
55 Solar Pro 4Upstage
119
$0.09 / $0.36
58 GPT-6 Luna (low reasoning)OpenAI
115
$0.10 / $0.50
59 GPT-5.6 Terra (extra-high reasoning)OpenAI
114
$2 / $12
60 GPT-6 Luna (no reasoning)OpenAI
113
$0.10 / $0.50
61 GPT-6 Luna (extra-high reasoning)OpenAI
112
$0.10 / $0.50
62 Qwen3.5-9B (no reasoning)Alibaba
110
$0.10 / $0.15
63 Qwen3-Omni-30B-A3B-ThinkingAlibaba
108
–
64 Qwen3-Omni-30B-A3B-InstructAlibaba
107
–
65 GPT-5.6 Terra (high reasoning)OpenAI
106
$2 / $12
66 Claude Sonnet 5.5 (extra-high reasoning)Anthropic
103
$2 / $10
67 Claude Sonnet 5.5 (high reasoning)Anthropic
99.8
$2 / $10
68 GPT-5.6 Terra (medium reasoning)OpenAI
99.7
$2 / $12
69 Llama 3.3 70B InstructMeta
97.7
$0.59 / $0.79
70 GPT-5.6 Terra (low reasoning)OpenAI
96.7
$2 / $12
71 Qwen3.5-Omni-PlusAlibaba
96.5
–
72 Claude Opus 5.5 (max reasoning)Anthropic
96.0
$4 / $20
73 DeepSeek-V4-Pro (0813, max reasoning)DeepSeek
95.9
$1.32 / $3.96
73 GPT-5.6 Terra (no reasoning)OpenAI
95.9
$2 / $12
75 Ministral 3 8BMistral AI
93.2
$0.15 / $0.15
76 Claude Sonnet 5.5 (medium reasoning)Anthropic
93.0
$2 / $10
76 DeepSeek-V4-Pro (0813, no reasoning)DeepSeek
93.0
$1.32 / $3.96
78 Claude Sonnet 5.5 (low reasoning)Anthropic
92.7
$2 / $10
79 GLM-5.3 (low reasoning)Z.ai
87.5
$1.40 / $4.40
80 Hy3Tencent
83.3
$0.14 / $0.58
80 Reka Flash 3Reka AI
83.3
$0.10 / $0.20
82 Gemma 4 26B A4B (no reasoning)Google
81.8
$0.10 / $0.30
83 Claude Opus 5.5 (extra-high reasoning)Anthropic
80.3
$4 / $20
84 Ministral 3 14BMistral AI
79.2
$0.20 / $0.20
85 GLM-5.3-FlashZ.ai
78.7
$0.15 / $0.50
86 GLM-5.3 (max reasoning)Z.ai
78.4
$1.40 / $4.40
87 Qwen3-Coder-NextAlibaba
77.3
$0.18 / $0.90
88 Magistral Small 1.2Mistral AI
76.0
–
89 Claude Opus 5.5 (low reasoning)Anthropic
75.6
$4 / $20
90 Claude Opus 5.5 (high reasoning)Anthropic
75.2
$4 / $20
91 Grok 4.7 (extra-high reasoning)xAI
74.3
$2 / $6
92 Claude Opus 5.5 (medium reasoning)Anthropic
74.0
$4 / $20
93 Grok 4.7 (high reasoning)xAI
73.7
$2 / $6
94 Claude Fable 5.1 (max reasoning)Anthropic
72.5
$10 / $50
95 GPT-6 Astra (max reasoning)OpenAI
71.7
$10 / $50
96 Qwen3.5-9BAlibaba
70.8
$0.10 / $0.15
97 GPT-6 Astra (extra-high reasoning)OpenAI
67.5
$10 / $50
98 Granite 4.2 8BIBM
64.9
$0.06 / $0.25
99 Grok 4.7 (low reasoning)xAI
63.4
$2 / $6
100 Mistral Large 4Mistral AI
60.2
$1.36 / $4.18
101 Claude Fable 5.1 (extra-high reasoning)Anthropic
60.1
$10 / $50
102 Command ACohere
59.5
$2.50 / $10
103 GPT-6 Astra (high reasoning)OpenAI
58.8
$10 / $50
104 GPT-6 Astra (low reasoning)OpenAI
56.2
$10 / $50
105 Qwen3.7-PlusAlibaba
56.1
$0.32 / $1.28
106 Qwen3.5-397B-A17B (no reasoning)Alibaba
55.6
$0.55 / $3.50
107 Qwen3.8-Flash-NextAlibaba
55.0
–
108 GPT-6 Astra (medium reasoning)OpenAI
54.8
$10 / $50
109 Qwen3.5-397B-A17BAlibaba
54.7
$0.55 / $3.50
110 Claude Fable 5.1 (medium reasoning)Anthropic
53.7
$10 / $50
110 Kimi K2.7 CodeMoonshot AI
53.7
$0.95 / $4
112 Gemma 4 26B A4BGoogle
53.1
$0.10 / $0.30
112 GPT-6.1 Sol (max reasoning)OpenAI
53.1
$2 / $10
114 Claude Fable 5.1 (low reasoning)Anthropic
51.8
$10 / $50
115 Claude Fable 5.1 (high reasoning)Anthropic
51.4
$10 / $50
116 GPT-6.1 Sol (extra-high reasoning)OpenAI
49.6
$2 / $10
117 GPT-6.1 Sol (high reasoning)OpenAI
48.5
$2 / $10
118 Gemma 4 31B (no reasoning)Google
47.6
$0.14 / $0.40
119 Qwen3.8-27B (low reasoning)Alibaba
46.4
$0.50 / $3
119 Llama 4 MaverickMeta
46.4
$0.27 / $0.85
121 MiMo-V2.6-FlashXiaomi
46.2
$0.14 / $0.28
122 Phi-4Microsoft
45.1
$0.07 / $0.14
123 GPT-6.1 Sol (low reasoning)OpenAI
45.0
$2 / $10
124 GPT-6.1 Sol (medium reasoning)OpenAI
44.5
$2 / $10
125 Qwen3.8-27B (medium reasoning)Alibaba
43.0
$0.50 / $3
126 Qwen3.8-27B (no reasoning)Alibaba
42.7
$0.50 / $3
126 MiMo-V2.6-ProXiaomi
42.7
$0.43 / $0.87
128 Kimi K3 (max reasoning)Moonshot AI
41.0
$3 / $15
129 Kimi K3 (low reasoning)Moonshot AI
39.9
$3 / $15
130 Qwen3.8-27B (extra-high reasoning)Alibaba
39.2
$0.50 / $3
131 Qwen3.8-Max (0902)Alibaba
36.8
$2 / $6
132 Qwen3.8-2.4T-A95BAlibaba
35.7
$2 / $6
133 Gemma 4 31BGoogle
35.4
$0.14 / $0.40
134 Gemma 4 E4B (no reasoning)Google
31.7
–
135 Gemma 4 E4BGoogle
28.0
–
136 Qwen3.5-4B (reasoning on)Alibaba
9.71
–
137 Qwen3.5-4B (no reasoning)Alibaba
9.02
–

Swipe the table sideways for more columns.

Ranks follow the score as shown, so equal numbers share a rank. Results as published by Artificial Analysis; we do not re-run them.

Progress over time

Output speed, plotted at each model's release date, higher is better. Each model counts once, at its best setting. Scores are today's, not launch day's. Hover or tap a mark for its model.

Best score so far

Every model at its release date. The line moves each time a new release beats every earlier one, and each new best is named where there is room.

0200400 2025Jul2026Jul Nova Micro 1.0gpt-oss-20bTrinity Large ThinkingGemini 3.5 Flash-LiteNemotron 3.5 Super VL

Best score by lab

Each lab's best so far, stepping whenever one of its releases beats its own earlier models. The 6 labs with the best scores today; 17 others not drawn.

NVIDIAGoogleDeepSeekArcee AIReflectionOpenAI

0100200300400500 2025Jul2026Jul NVIDIAGoogleDeepSeekArcee AIReflectionOpenAI

What it measures

Median tokens a second while the model writes its answer, measured by Artificial Analysis on the model's usual API over the past weeks.

What it does not measure

Not time spent thinking before the answer starts, or speed on other hosts.

Source

Artificial Analysis · Methodology

285 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: 270
  • A different snapshot or variant from the model we list: 14
  • An unusual combination of settings: 1

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