Which model leads ScreenSpot?

Qwen3 VL 32B Instruct leads the ScreenSpot board of 12 measured models on the Vector Wire tracker at 95.8. Every score is source-attributed with a verification tier.

ScreenSpot Leaderboard — Model Scores

Refresh Models —Deals —Regulatory —Sources —

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

The field · 12 measured

87.195.8

12 models measured, most on the official card harness. Qwen3 VL 32B Instruct tops the board at 95.8.

12 measured·12 new this week·Lifecycle Updated Aug 27, 2026Basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 12Bars show 87.1–95.8· ◆ solid = first-party or better · outlined = arm's-length

01Qwen3 VL 32B InstructNEWAlibaba95.8$0.16/M$0.64/M$0.004

02Qwen3 VL 32B ThinkingNEWAlibaba95.7$0.16/M$0.64/M$0.004

03Qwen3 VL 235B A22B ThinkingNEWAlibaba95.4$0.40/M$4.0/M$0.023

04Qwen3 VL 235B A22B InstructNEWAlibaba95.4$0.40/M$1.6/M$0.010

05Qwen3 VL 30B A3B InstructNEWAlibaba94.7$0.20/M$0.80/M$0.005

06Qwen3 VL 30B A3B ThinkingNEWAlibaba94.7$0.20/M$2.4/M$0.014

07Qwen3 VL 8B InstructNEWAlibaba94.4$0.18/M$0.70/M$0.005

08Qwen3 VL 4B InstructNEWAlibaba94———

09Qwen3 VL 8B ThinkingNEWAlibaba93.6$0.18/M$2.1/M$0.012

10Qwen3 VL 4B ThinkingNEWAlibaba92.9———

11Qwen2.5 VL 32B InstructNEW+1 altAlibaba88.5———

12Qwen2.5 VL 72B InstructNEWAlibaba87.1———

◆ 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.