New API

API Provider Performance Test Results

Host Addressai.081007.xyz
Supported Models11 models
Supported Models
11 models
Model NameTest CountAverage Output SpeedAverage First Token Latency
Qwen3-235B-A22B
10
40.69 t/s
7.58s
gpt-oss-20b
10
229.65 t/s
1.77s
DeepSeek-V3-0324
5
49.39 t/s
17.04s
doubao-1.6
5
46.33 t/s
19.62s
gemini-2.5-flash
5
208.59 t/s
8.73s
gemini-2.5-pro
5
16.08 t/s
14.26s
GLM-4.5-Air
5
64.75 t/s
3.72s
kimi-k2-instruct
5
41.95 t/s
1.62s
command-a-03-2025
5
120.65 t/s
0.52s
ST-dolphin-mistral-24b-未经审查
5
27.10 t/s
1.83s
DeepSeek-R1-0528
5
34.88 t/s
13.67s
Total Tests
65

Total Tests

Average Tokens/Second
88.49 t/s

0.00 - 278.43 t/s

Average First Token Latency
7.67s

0.42 - 72.73s

Last Test Time
8/13/2025, 2:30:25 PM

Last Test Time

Top Models Performance
1

gpt-oss-20b

Test Count: 10

229.65 t/s

Average Speed

2

gemini-2.5-flash

Test Count: 5

208.59 t/s

Average Speed

3

command-a-03-2025

Test Count: 5

120.65 t/s

Average Speed

4

GLM-4.5-Air

Test Count: 5

64.75 t/s

Average Speed

5

DeepSeek-V3-0324

Test Count: 5

49.39 t/s

Average Speed

Recent Test Records
10 records
Test TimeModelAverage Output SpeedAverage First Token LatencyTotal Tokens
8/13/2025, 2:30:25 PM
Qwen3-235B-A22B
31.46 t/s
1.49s
2958
8/13/2025, 2:26:57 PM
gemini-2.5-flash
208.59 t/s
8.73s
4660
8/13/2025, 2:24:48 PM
gemini-2.5-pro
16.08 t/s
14.26s
1139
8/13/2025, 2:05:45 PM
kimi-k2-instruct
41.95 t/s
1.62s
3619
8/13/2025, 1:59:03 PM
Qwen3-235B-A22B
49.92 t/s
13.66s
4712
8/13/2025, 1:55:09 PM
DeepSeek-R1-0528
34.88 t/s
13.67s
4582
8/13/2025, 1:52:32 PM
DeepSeek-V3-0324
49.39 t/s
17.04s
2489
8/13/2025, 1:51:42 PM
gpt-oss-20b
218.16 t/s
1.74s
4371
8/13/2025, 1:45:03 PM
ST-dolphin-mistral-24b-未经审查
27.10 t/s
1.83s
2040
8/13/2025, 1:42:23 PM
GLM-4.5-Air
64.75 t/s
3.72s
5391