Real-world LLM speed leaderboard

Anonymous real-world speed results from TOKRACE, annotated by sample confidence. Current 11 models · 581 runs

ModelMedian tok/sTTFTPeak
🥇kimi-for-coding-highspeed
api.kimi.com· 3 runsLow sample
224
avg 231
1.27s
467
🥈kimi-for-coding
api.kimi.com· 129 runsStable sample
184
avg 225
1.49s
900
🥉step-3.7-flash
api.stepfun.com· 69 runsStable sample
170
avg 178
1.99s
2,067
4deepseek-v4-flash
api.deepseek.com· 88 runsStable sample
142
avg 132
0.80s
638
5MiniMax-M3
api.minimaxi.com· 6 runsLow sample
100
avg 97
1.52s
395
6mimo-v2.5
api.xiaomimimo.com· 45 runsUsable signal
86
avg 88
4.38s
514
7deepseek-v4-pro
api.deepseek.com· 55 runsStable sample
85
avg 86
1.29s
396
8glm-5.1
open.bigmodel.cn· 52 runsStable sample
53
avg 61
4.64s
594
9glm-5.2
open.bigmodel.cn· 81 runsStable sample
51
avg 60
3.90s
846
10mimo-v2.5-pro
api.xiaomimimo.com· 41 runsUsable signal
43
avg 48
1.98s
225
11Kimi K3
api.kimi.com· 12 runsUsable signal
36
avg 46
3.09s
157

Speed vs cost

Full pricing table →

Median output speed against output price. Top-left is best (fast & cheap); the dashed line marks the Pareto frontier — models not beaten on both speed and price.

062123185247$0.30$1$3$10Output $/1M (log, ← cheaper)Median tok/s (faster ↑)↖ fast & cheap = bestkimi-for-coding-highspeed · api.kimi.com speed 224 tok/s · output $4/1M · Pareto frontierkimi-for-coding-highspeedkimi-for-coding · api.kimi.com speed 184 tok/s · output $4/1Mstep-3.7-flash · api.stepfun.com speed 170 tok/s · output $1/1M · Pareto frontierstep-3.7-flashdeepseek-v4-flash · api.deepseek.com speed 142 tok/s · output $0.28/1M · Pareto frontierdeepseek-v4-flashMiniMax-M3 · api.minimaxi.com speed 100 tok/s · output $1/1Mmimo-v2.5 · api.xiaomimimo.com speed 86 tok/s · output $0.28/1M · Pareto frontiermimo-v2.5deepseek-v4-pro · api.deepseek.com speed 85 tok/s · output $0.87/1Mglm-5.1 · open.bigmodel.cn speed 53 tok/s · output $4/1Mglm-5.2 · open.bigmodel.cn speed 51 tok/s · output $4/1Mmimo-v2.5-pro · api.xiaomimimo.com speed 43 tok/s · output $0.28/1MKimi K3 · api.kimi.com speed 36 tok/s · output $14/1M

· Cost uses each model's output price per 1M tokens (CNY converted to USD); speed is the median from the leaderboard above.

Popular comparisons

Head-to-head speed pages for the top models on the board.

kimi-for-coding vs kimi-for-coding-highspeedkimi-for-coding-highspeed vs step-3.7-flashdeepseek-v4-flash vs kimi-for-coding-highspeedkimi-for-coding-highspeed vs MiniMax-M3kimi-for-coding-highspeed vs mimo-v2.5deepseek-v4-pro vs kimi-for-coding-highspeedglm-5.1 vs kimi-for-coding-highspeedkimi-for-coding vs step-3.7-flashdeepseek-v4-flash vs kimi-for-codingkimi-for-coding vs MiniMax-M3kimi-for-coding vs mimo-v2.5deepseek-v4-pro vs kimi-for-codingglm-5.1 vs kimi-for-codingdeepseek-v4-flash vs step-3.7-flashMiniMax-M3 vs step-3.7-flashmimo-v2.5 vs step-3.7-flashdeepseek-v4-pro vs step-3.7-flashglm-5.1 vs step-3.7-flashdeepseek-v4-flash vs MiniMax-M3deepseek-v4-flash vs mimo-v2.5deepseek-v4-flash vs deepseek-v4-prodeepseek-v4-flash vs glm-5.1mimo-v2.5 vs MiniMax-M3deepseek-v4-pro vs MiniMax-M3glm-5.1 vs MiniMax-M3deepseek-v4-pro vs mimo-v2.5glm-5.1 vs mimo-v2.5deepseek-v4-pro vs glm-5.1

· Data comes from voluntary anonymous sharing and contains speed metrics only, not prompts or API keys · Updates every 5 minutes

· Speed is affected by network, time of day and provider load; different endpoints for the same model are tracked separately