Which model leads charxiv_rq_with_python?

charxiv_rq_with_python Leaderboard — Model Scores

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

The field · 9 measured

78.790

9 models measured, most on the third-party eval harness. GPT-5.4 tops the board at 90.

9 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 9Bars show 78.7–90 · ◆ solid = first-party or better · outlined = arm's-length

01GPT-5.4OpenAI90$2.5/M$15.0/M$0.097

02Gemini 3.1 Pro+2 altsGoogle89.9$2.0/M$12.0/M$0.078

03Claude Fable 5Anthropic89.4$10.0/M$50.0/M$0.336

04GPT-5.6 SolOpenAI87.8$4.0/M$20.0/M$0.137

05Kimi K2.6+1 altMoonshot86.7$0.95/M$4.0/M$0.029

06Claude Opus 4.6Anthropic84.7$5.0/M$25.0/M$0.177

07Inkling+1 altThinking Machines83.4VENDOR$0.95/M$4.0/M$0.030

08Inkling-SmallThinking Machines81.3VENDOR$0.30/M$1.2/M$0.009

09Kimi K2.5+1 altMoonshot78.7$0.45/M$2.3/M$0.017

◆ 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: 9 models scored · 0 independently verified · 5 vendor cross-reference · 4 vendor-reported · 0 with source disagreement. How these tiers are assigned