Output Speed

Artificial Analysis · 50 models

Rank	Model	Score
1	Celeris-1	1,638.45
2	Nemotron 3.5 Super VL	450.23
3	Gemini 3.5 Flash-Lite	423.32
4	Gemini 2.5 Flash-Lite (Reasoning)	407.39
5	Gemini 3.1 Flash-Lite	390.77
6	Beam 1.0	356.77
7	Muse Spark 1.3 (Xhigh)	334.55
8	Ling 3.0 Flash	320.7
9	K2 Horizon 375B A23B	312.58
10	Gemini 2.5 Flash-Lite (Non-reasoning)	311.72
11	Nemotron 3.5 Lightning	297.85
12	GPT-5.4 mini (Xhigh)	293.34
13	Nova Micro	277.45
14	GPT-4.1 nano	275.02
15	Muse Spark 1.1 (Xhigh)	273.05
16	NVIDIA Nemotron Nano 9B V2 (Reasoning)	264.18
17	DeepSeek V4.1 Flash (Non-reasoning)	262.98
18	gpt-oss-120b (Low)	257.83
19	NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)	257.61
20	Muse Spark 1.2 (Xhigh)	256.86
21	o3-mini (High)	255.87
22	Qwen3 Coder 30B A3B Instruct	242.72
23	Qwen3.5 Omni Flash	240.9
24	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)	237.28
25	GPT-5.4 mini (Non-reasoning)	237.18
26	Claude Haiku 5.5 (Max)	235.38
27	Inkling Small	232.47
28	Nemotron 3 Nano Omni 30B A3B Reasoning	230.43
29	GPT-5.4 mini (Medium)	230.29
30	o3-mini	228.98
31	DeepSeek V4.1 Flash 1017 (High)	218.3
32	Ling 3.1 Flash	217.73
33	Gemini 2.5 Flash (Reasoning)	211.87
34	gpt-oss-120b (High)	211.68
35	DeepSeek V4.1 Flash (Max)	211.52
36	Inkling (Xhigh)	211.28
37	GPT-5 nano (Minimal)	209.9
38	GLM-5.2 (Non-reasoning)	209.8
39	gpt-oss-20b (Low)	209.03
40	Gemini 2.5 Flash (Non-reasoning)	207.94
41	Gemini 3 Flash Preview (Reasoning)	203.29
42	Qwen3.7 Max	202.54
43	NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)	199.17
44	Qwen3 30B A3B 2507 Instruct	198.84
45	Gemini 3.6 Flash (High)	197.55
46	Muse Wonder (Max)	195.86
47	Claude Haiku 5.5 (Xhigh)	193.63
48	Gemini 3.5 Flash (Medium)	192.73
49	GPT-5 Codex (High)	192.67
50	Nova 2.0 Lite (Medium)	192.12