Which model leads theagentcompany?

Q: Which model leads theagentcompany?

Claude Sonnet 4.5 tops the board at 41

theagentcompany Leaderboard — Model Scores

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

The field · 7 measured

3041

7 models measured, most on the third-party eval harness. Claude Sonnet 4.5 tops the board at 41.

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 30–41 · ◆ solid = first-party or better · outlined = arm's-length

01Claude Sonnet 4.5Anthropic41$3.0/M$15.0/M$0.220

02Gemini 2.5 ProGoogle39.3$1.3/M$10.0/M$0.143

03Claude Sonnet 4Anthropic37$3.0/M$15.0/M$0.243

04MiniMax M2MiniMax36$0.30/M$1.2/M$0.021

05GLM-4.6Z.ai35$0.57/M$2.2/M$0.040

06DeepSeek-V3.2DeepSeek34$0.28/M$0.42/M$0.010

07Kimi K2 InstructMoonshot30$0.57/M$2.3/M$0.048

◆ 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 · 6 vendor cross-reference · 1 vendor-reported · 0 with source disagreement. How these tiers are assigned