Which model leads IFBench (prompt loose)?

Q: Which model leads IFBench (prompt loose)?

Nemotron 3 Ultra 550B A55B tops the board at 81.7

IFBench (prompt loose) Leaderboard — Model Scores

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

The field · 5 measured

73.781.7

5 models measured, most on the official card harness. Nemotron 3 Ultra 550B A55B tops the board at 81.7.

5 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 5Bars show 73.7–81.7 · ◆ solid = first-party or better · outlined = arm's-length

01Nemotron 3 Ultra 550B A55BNVIDIA81.7$0.60/M$2.5/M$0.019

02Qwen3.5 397B A17BAlibaba78.2$0.60/M$3.6/M$0.027

03GLM-5.1Z.ai76.6$1.4/M$4.4/M$0.038

04MiniMax M2.7MiniMax74.6$0.21/M$0.84/M$0.007

05Kimi K2.6Moonshot73.7$0.95/M$4.0/M$0.034

◆ 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: 5 models scored · 0 independently verified · 4 vendor cross-reference · 1 vendor-reported · 0 with source disagreement. How these tiers are assigned