ModelSWE-bench VerifiedPer taskPriceSpeedSupportsContext
83.5max
—$5 / $25521M
GLM 5.2Z.aiOpen
78.7max
$1.28$1.40 / $4.40941M
77.6max
—$1.32 / $3.96—1M
Qwen3.7 MaxAlibaba (Qwen)
77.3max
—$1.48 / $4.43—1M
76.9high
—$2.50 / $15—1.05M
76.7unstated
—$5 / $25—200K
Kimi K2.6Moonshot AIOpen
76.7unstated
—$0.95 / $4—256K
Qwen3.6 Max PreviewAlibaba (Qwen)
76.7unstated
—$1.30 / $7.80—262K
75.6unstated
—$2 / $12—1M
75.4unstated
—$0.50 / $3—1M
75.2unstated
—$3 / $15—1M
74.8high
—$1.75 / $14—400K
GLM 5.1Z.aiOpen
74.2unstated
—$1.40 / $4.40—203K
Kimi K2.5Moonshot AIOpen
73.8unstated
—$0.60 / $3—256K
73.8high
—$1.75 / $14—400K
73.6high
—$1.25 / $10—400K
GLM 5Z.aiOpen
72.1unstated
—$1 / $3.20—203K
71.3unstated
—$3 / $15—1M
64.7medium
—$0.25 / $2—400K
62.3medium
—$2 / $8—200K
Qwen3.6 PlusAlibaba (Qwen)
57.9unstated
—$0.50 / $3—1M
57.6unstated
—$1.25 / $10—1M
48.5
—$2 / $8—1M
GPT-4oOpenAI
31unstated
—$2.50 / $10—128K