Which model leads FRAMES (Acc.)?

Q: Which model leads FRAMES (Acc.)?

DeepSeek-R1 tops the board at 82.5

FRAMES (Acc.) Leaderboard — 10 Models Ranked

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

The field · 10 measured

65.482.5

10 models measured, most on the official card harness. DeepSeek-R1 tops the board at 82.5.

10 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 10Bars show 65.4–82.5 · ◆ solid = first-party or better · outlined = arm's-length

01DeepSeek-R1+6 altsDeepSeek82.5$2.0/M$4.0/M$0.036

02DeepSeek-R1-ZeroDeepSeek82.3———

03GPT-4o+5 altsOpenAI80.5$5.0/M$15.0/M$0.124

04O1 Mini+4 altsOpenAI76.9———

05DeepSeek-V3+5 altsDeepSeek73.3$0.24/M$0.90/M$0.008

06Claude 3.5 Sonnet+5 altsAnthropic72.5$3.0/M$15.0/M$0.124

07Llama 3.1 Instruct 405BMeta70———

08Qwen2.5 Instruct 72BAlibaba69.8$0.47/M$0.49/M$0.007

09DeepSeek-V2DeepSeek66.9———

10DeepSeek-V2.5DeepSeek65.4———

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