Which model leads JHumanEval+?

Granite-4.0-1B leads the JHumanEval+ board of 6 measured models on the Vector Wire tracker at 51.22. Every score is source-attributed with a verification tier.

JHumanEval+ Leaderboard — Model Scores

Refresh Models —Deals —Regulatory —Sources —

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

The field · 6 measured

17.6851.22

6 models measured, most on the third-party eval harness. Granite-4.0-1B tops the board at 51.22.

6 measured·Lifecycle Updated Sep 2, 2026Basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 6Bars show 17.68–51.22· ◆ solid = first-party or better · outlined = arm's-length

01Granite-4.0-1BIBM51.22———

02LFM2.5-1.2B-JP-202606Liquid AI49.39———

03Qwen3-1.7B (Instruct)Alibaba47.56———

04LFM2.5-1.2B-InstructLiquid AI28.66———

05Gemma-3-1B-itGoogle25———

06Llama-3.2-1B-InstructMeta17.68$0.03/M$0.20/M$0.006

◆ 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.