Which model leads Finance Agent?

Q: Which model leads Finance Agent?

Claude Opus 4.7 tops the board at 64.4

Finance Agent Leaderboard — 11 Models Ranked

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

The field · 11 measured

53.764.4

11 models measured, most on the official card harness. Claude Opus 4.7 tops the board at 64.4.

11 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 11Bars show 53.7–64.4 · ◆ solid = first-party or better · outlined = arm's-length

01Claude Opus 4.7+4 altsAnthropic64.4$5.0/M$25.0/M$0.233

02Claude Sonnet 4.6Anthropic63.3$3.0/M$15.0/M$0.142

03GPT-5.4 Pro+1 altOpenAI61.5$30.0/M$180.0/M$1.707

04Claude Opus 4.6+3 altsAnthropic60.7$5.0/M$25.0/M$0.247

05GPT-5.5+1 altOpenAI60$5.0/M$30.0/M$0.292

06Gemini 3.1 Pro+2 altsGoogle59.7$2.0/M$12.0/M$0.117

07Gemini 3.5 Flash+2 altsGoogle57.9$1.5/M$9.0/M$0.091

08GPT-5.4+2 altsOpenAI56$2.5/M$15.0/M$0.156

09Claude Opus 4.5Anthropic55.2$5.0/M$25.0/M$0.272

10Claude Opus 4.8+1 altAnthropic53.9$5.0/M$25.0/M$0.278

11Nemotron 3 Ultra 550B A55BNVIDIA53.7$0.60/M$2.5/M$0.029

◆ 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: 11 models scored · 0 independently verified · 2 vendor cross-reference · 9 vendor-reported · 2 with source disagreement. How these tiers are assigned