Which model leads AgentWorldBench - Search?

Q: Which model leads AgentWorldBench - Search?

Qwen AgentWorld 397B A17B tops the board at 37.82

AgentWorldBench - Search Leaderboard

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

The field · 14 measured

21.9437.82

14 models measured, most on the third-party eval harness. Qwen AgentWorld 397B A17B tops the board at 37.82.

14 measured·Lifecycle Updated Oct 8, 2026Score basis

#◆Vendor$/M in$/M out

Leader — ignites whiteField — fill cools with scoreIndependently verified only0 of 14Bars show 21.94–37.82 · ◆ solid = first-party or better · outlined = arm's-length

01Qwen AgentWorld 397B A17BAlibaba37.82———

02GPT-5.4OpenAI37.26$2.5/M$15.0/M$0.235

03Qwen AgentWorld 35B A3BAlibaba36.69———

04Claude Opus 4.8Anthropic35.14$5.0/M$25.0/M$0.427

05Qwen3.5 397B A17BAlibaba30.81$0.60/M$3.6/M$0.068

06Gemini 3.1 ProGoogle30.21$2.0/M$12.0/M$0.232

07Claude Opus 4.6Anthropic29.3$5.0/M$25.0/M$0.512

08Claude Sonnet 4.6Anthropic28.79$3.0/M$15.0/M$0.313

09DeepSeek-V4-ProDeepSeek27.61$0.43/M$0.87/M$0.024

10Kimi K2.6Moonshot27.48$0.95/M$4.0/M$0.090

11MiniMax M2.7MiniMax27.3$0.21/M$0.84/M$0.019

12Qwen3.5 35B A3BAlibaba25.98$0.25/M$2.0/M$0.043

13GLM-5.1Z.ai22.46$1.4/M$4.4/M$0.129

14Qwen3.6 PlusAlibaba21.94$0.50/M$3.0/M$0.080

◆ 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: 14 models scored · 0 independently verified · 9 vendor cross-reference · 5 vendor-reported · 0 with source disagreement. How these tiers are assigned