Which model leads SpreadsheetBench 2?

SpreadsheetBench 2 Leaderboard — Model Scores

Pro Mode — raw scores, one benchmark under the lensthird-party eval harness

The field · 7 measured

21.8134.8

7 models measured, most on the third-party eval harness. Kimi K3 tops the board at 34.8.

7 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 7Bars show 21.81–34.8 · ◆ solid = first-party or better · outlined = arm's-length

01Kimi K3+1 altMoonshot34.8VENDOR$0.58/M$12.3/M$0.185

02Claude Fable 5Anthropic34.7$10.0/M$50.0/M$0.865

03GPT-5.6 SolOpenAI32.4$4.0/M$20.0/M$0.370

04Claude Opus 4.8Anthropic31.6$5.0/M$25.0/M$0.475

05GPT-5.5OpenAI29.1$5.0/M$30.0/M$0.601

06GLM-5.2Z.ai28.1$1.4/M$4.4/M$0.103

07Ling-3.0-flash-FinInclusionAI21.81VENDOR$0.04/M$0.12/M$0.004

◆ 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: 7 models scored · 0 independently verified · 5 vendor cross-reference · 2 vendor-reported · 0 with source disagreement. How these tiers are assigned