Which model leads WorldVQA?

Q: Which model leads WorldVQA?

Qwen3.7 Plus Preview tops the board at 61.1

WorldVQA Leaderboard — 12 Models Ranked

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

The field · 12 measured

23.561.1

12 models measured, most on the official card harness. Qwen3.7 Plus Preview tops the board at 61.1.

12 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 12Bars show 23.5–61.1 · ◆ solid = first-party or better · outlined = arm's-length

01Qwen3.7 Plus PreviewAlibaba61.1$0.40/M$1.6/M$0.016

02Seed 2.1 Pro Preview+1 altByteDance53———

03Kimi K3Moonshot51$0.58/M$12.3/M$0.126

04Seed2.1ByteDance48.6———

05Gemini 3 ProGoogle47.43P$2.0/M$12.0/M$0.148

06Kimi K2.5+1 altMoonshot46.3$0.45/M$2.3/M$0.029

07Gemini 3.1 ProGoogle44.3$2.0/M$12.0/M$0.158

08Claude Opus 4.5Anthropic36.83P$5.0/M$25.0/M$0.408

09Claude Opus 4.7Anthropic35.9$5.0/M$25.0/M$0.418

10GPT-5.5OpenAI34.6$5.0/M$30.0/M$0.506

11GPT-5.2OpenAI283P$1.8/M$14.0/M$0.281

12Qwen3 VL 235B A22B ReasoningAlibaba23.53P$0.40/M$4.0/M$0.094

◆ 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: 12 models scored · 0 independently verified · 7 vendor cross-reference · 5 vendor-reported · 0 with source disagreement. How these tiers are assigned