Which model leads ProgramBench (Almost Solved)?

GPT-5.6 Sol leads the ProgramBench (Almost Solved) board of 6 measured models on the Vector Wire tracker at 23. Every score is source-attributed with a verification tier.

ProgramBench (Almost Solved) Leaderboard — Model Scores

Refresh Models —Deals —Regulatory —Sources —

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

The field · 6 measured

9.523

6 models measured, most on the third-party eval harness. GPT-5.6 Sol tops the board at 23.

6 measured·6 new this week·Lifecycle Updated Sep 2, 2026Basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 6Bars show 9.5–23· ◆ solid = first-party or better · outlined = arm's-length

01GPT-5.6 SolNEW+1 altOpenAI23$4.0/M$20.0/M$0.522

02GLM-5.3NEW+1 altZ.ai19$1.4/M$4.4/M$0.153

03Kimi K3NEW+1 altMoonshot17.5$2.9/M$14.3/M$0.489

04Opus 4.8NEW+1 altAnthropic15.5$5.0/M$25.0/M$0.968

05Qwen3.8-MaxNEW+1 altAlibaba10.5$2.0/M$6.0/M$0.381

06GLM-5.2NEW+1 altZ.ai9.5$1.4/M$4.4/M$0.305

◆ 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.