Rankings

Model rankings

The most-used models on the Flow AI network over the last 7 days — with the live clearing price and how far below published rates they cleared. Updated every few minutes.

#ModelRequests (7d)Avg latencyPrice $/M outSavings
1byo:Qwen 3.8 Flash
58,656
9551 ms——
2byo:MiniMax-M3
46,839
7055 ms——
3qwen3-embedding-4b
37,333
3042 ms——
4byo:GLM 4.6 (plan)
19,732
4776 ms——
5byo:Grok 4.6
7,134
26436 ms——
6byo:GLM 5.2
6,680
3922 ms——
7byo:GLM 4.6V (plan)
5,916
10956 ms——
8byo:Qwen 3.8 Max
5,410
15818 ms——
9byo:MiniMax-M2.1
5,383
5985 ms——
10byo:MiniMax-M2.7
4,674
31805 ms——
11byo:MiniMax-M2.5-highspeed
4,123
25825 ms——
12byo:GLM 4.5 (plan)
3,384
6870 ms——
13byo:MiniMax-M2.7-highspeed
2,979
33963 ms——
14byo:Kimi For Coding
2,702
9340 ms——
15byo:MiniMax-M2.1-highspeed
2,585
10068 ms——
16byo:Kimi K3 (plan)
1,998
27728 ms——
17byo:MiniMax-M2.5
1,916
8799 ms——
18byo:MiniMax-M3.1-Flash
1,750
10086 ms——
19byo:MiniMax-M2
1,585
10980 ms——
20byo:GLM 4.5V (plan)
1,520
17611 ms——
21byo:Qwen 3.7 Plus
709
24279 ms——
22byo:MiMo Subscription
665
29647 ms——
23byo:GLM 5.3 Flash (plan)
309
28957 ms——
24gemini-2.5-flash
239
6704 ms$2.5000—
25byo:GLM 4.5-air (plan)
238
29430 ms——
26kimi-k2.6-search
174
42327 ms$2.5000—
27gpt-4o-mini-online
174
8724 ms$0.6000—
28sonar
173
3896 ms$1.0000—
29byo:GLM Coding Max
172
7268 ms——
30byo:GLM 5.1
145
5792 ms——
31qwen3-27b
53
6118 ms$0.8000—
32deepseek-v4-flash
39
2050 ms$0.6000—
33or:nvidia/nemotron-3.5-super-vl:free
19
5228 ms$0.0000—
34or:nvidia/nemotron-3-super-120b-a12b:free
15
11003 ms$0.0000—
35or:nvidia/nemotron-3-ultra-550b-a55b:free
14
4901 ms$0.0000—
36mistral-smallself-host
7
5695 ms$0.0800—
37qwen3-8bself-host
7
12346 ms$0.1600—
38nemotron-3-ultra
6
21713 ms$0.0000—
39or:cohere/north-mini-code:free
6
3312 ms$0.0000—
40or:liquid/lfm-2.5-2.6b:free
2
1769 ms$0.0000—

See all models & live prices or how Flow AI compares to OpenRouter.