Which model leads BrowseComp (w/ Ctx)?

Q: Which model leads BrowseComp (w/ Ctx)?

GPT-5.6 Sol tops the board at 89.4

BrowseComp (w/ Ctx) Leaderboard — Model Scores

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

The field · 7 measured

74.989.4

7 models measured, most on the third-party eval harness. GPT-5.6 Sol tops the board at 89.4.

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 74.9–89.4 · ◆ solid = first-party or better · outlined = arm's-length

01GPT-5.6 SolOpenAI89.4$4.0/M$20.0/M$0.134

02Claude Fable 5Anthropic88$10.0/M$50.0/M$0.341

03Gemini 3.1 ProGoogle85.9$2.0/M$12.0/M$0.081

04DeepSeek-V4-ProDeepSeek83.4$0.43/M$0.87/M$0.008

05Kimi K2.6Moonshot83.2$0.95/M$4.0/M$0.030

06InklingThinking Machines77.1$0.95/M$4.0/M$0.032

07Kimi K2.5Moonshot74.9$0.45/M$2.3/M$0.018

◆ 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