Which model leads ScreenSpot-v2 Web?

InternVL 3.5 4B leads the ScreenSpot-v2 Web board of 7 measured models on the Vector Wire tracker at 82.6. Every score is source-attributed with a verification tier.

ScreenSpot-v2 Web Leaderboard — Model Scores

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

The field · 7 measured

2.582.6

7 models measured, most on the official card harness. InternVL 3.5 4B tops the board at 82.6.

7 measured·6 new this week·Lifecycle Updated Sep 12, 2026Basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 7Bars show 2.5–82.6· ◆ solid = first-party or better · outlined = arm's-length

01InternVL 3.5 4BNEWOpenGVLab82.6———

02LFM2.5-VL-3BLiquid AI82.2———

03InternVL 3.5 2BNEWOpenGVLab79.9———

04Qwen3.5-2BNEWAlibaba65.9———

05gemma-4-E4B-itNEWGoogle47.6$0.02/M$0.10/M$0.001

06gemma-4-E2B-itNEWGoogle22.4———

07LFM2-VL-3BNEWLiquid AI2.5———

◆ 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.