Model latency
Time to first token and output throughput from the last 24 hours, grouped by Freebuff model.
Updated just now
Median TTFT
2.1s
1M samples in 24h
P95 TTFT
6.7s
Slower tail across all models
Output speed
111 tok/s
Output tokens per second after TTFT
Lowest Median
426ms
MiniMax M2.7 · 1,294 samples
Models
7
Reporting TTFT samples
By model
Sorted by median TTFT. Bars show P95 latency.
Last 24 hours
MiniMax M2.7
minimax/minimax-m2.7
Median
426ms
P95
3.8s
Tok/s
256
1,294 sampleshourly median
Kimi K2.6
moonshotai/kimi-k2.6
Median
1.4s
P95
5.2s
Tok/s
140
7,867 sampleshourly median
MiniMax M3
minimax/minimax-m3
Median
1.6s
P95
12s
Tok/s
104
129K sampleshourly median
DeepSeek V4 Flash
deepseek/deepseek-v4-flash
Median
2.1s
P95
3.5s
Tok/s
124
673K sampleshourly median
DeepSeek V4 Pro
deepseek/deepseek-v4-pro
Median
2.3s
P95
5.0s
Tok/s
97
110K sampleshourly median
MiMo 2.5 Pro
mimo/mimo-v2.5-pro
Median
4.7s
P95
15s
Tok/s
50
15K sampleshourly median
MiMo 2.5
mimo/mimo-v2.5
Median
5.4s
P95
14s
Tok/s
95
94K sampleshourly median