Which model leads MCPMark?

Q: Which model leads MCPMark?

Kimi K2.7 Code tops the board at 81.1

MCPMark Leaderboard — Model Scores

Pro Mode — raw scores, one benchmark under the lensofficial card harness

The field · 17 measured

14.281.1

17 models measured, most on the official card harness. Kimi K2.7 Code tops the board at 81.1.

17 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 17Bars show 14.2–81.1 · ◆ solid = first-party or better · outlined = arm's-length

01Kimi K2.7 CodeMoonshot81.1$1.9/M$8.1/M$0.062

02GPT-5.4OpenAI62.53P$2.5/M$15.0/M$0.140

03Qwen3.7 MaxAlibaba60.8$2.5/M$7.5/M$0.082

04Qwen3.7 Plus PreviewAlibaba58.7$0.40/M$1.6/M$0.017

05Claude Opus 4.6Anthropic56.73P$5.0/M$25.0/M$0.265

06Gemini 3.1 Pro+1 altGoogle55.93P$2.0/M$12.0/M$0.125

07Kimi K2.6+1 altMoonshot55.9$0.95/M$4.0/M$0.044

08Qwen3.6 PlusAlibaba48.2$0.50/M$3.0/M$0.036

09Qwen3.5 397B A17BAlibaba46.1$0.60/M$3.6/M$0.046

10DeepSeek-V3.2DeepSeek38$0.28/M$0.42/M$0.009

11Qwen3.6 35B A3B+1 altAlibaba37$0.38/M$2.3/M$0.035

12Qwen3.5 27BAlibaba36.33P$0.30/M$2.4/M$0.037

13Qwen3 Max (Reasoning)Alibaba33.5$1.2/M$6.0/M$0.107

14Kimi K2.5Moonshot29.53P$0.45/M$2.3/M$0.046

15Qwen3.5 35B A3BAlibaba273P$0.25/M$2.0/M$0.042

16Gemma 4 31BGoogle18.13P$0.17/M$0.40/M$0.016

17Gemma 4 26B A4BGoogle14.23P$0.07/M$0.34/M$0.014

◆ solid = independently verified · aggregator attested · vendor attributed — outlined = cross-reference · source-attributed. Dimmed ◆ = confidence inferred. “+N alts” = alternate disclosures in the drawer with why-demoted. Prices per million tokens; $/point = input price ÷ score. Every row opens its model dossier.

Verification: 17 models scored · 0 independently verified · 5 vendor cross-reference · 12 vendor-reported · 0 with source disagreement. How these tiers are assigned