Which model leads AgentWorldBench - SWE?

Q: Which model leads AgentWorldBench - SWE?

Qwen AgentWorld 397B A17B tops the board at 68.49

AgentWorldBench - SWE Leaderboard — Model Scores

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

The field · 14 measured

37.4468.49

14 models measured, most on the third-party eval harness. Qwen AgentWorld 397B A17B tops the board at 68.49.

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 37.44–68.49 · ◆ solid = first-party or better · outlined = arm's-length

01Qwen AgentWorld 397B A17BAlibaba68.49———

02GPT-5.4OpenAI66.29$2.5/M$15.0/M$0.132

03Qwen AgentWorld 35B A3BAlibaba65.63———

04Claude Opus 4.6Anthropic64.55$5.0/M$25.0/M$0.232

05Claude Sonnet 4.6Anthropic64.52$3.0/M$15.0/M$0.139

06Qwen3.5 397B A17BAlibaba64.44$0.60/M$3.6/M$0.033

07Claude Opus 4.8Anthropic64.1$5.0/M$25.0/M$0.234

08DeepSeek-V4-ProDeepSeek59.44$0.43/M$0.87/M$0.011

09Qwen3.6 PlusAlibaba59.08$0.50/M$3.0/M$0.030

10Gemini 3.1 ProGoogle59.07$2.0/M$12.0/M$0.119

11Kimi K2.6Moonshot58.77$0.95/M$4.0/M$0.042

12GLM-5.1Z.ai52.07$1.4/M$4.4/M$0.056

13Qwen3.5 35B A3BAlibaba47.58$0.25/M$2.0/M$0.024

14MiniMax M2.7MiniMax37.44$0.21/M$0.84/M$0.014

◆ 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