List your company

SQL Generation on BacktestBench synthetic (test)

100Execution Correctness Rate (ECR)Qwen3 235B
90.806493.193295.5897.9668May 18, 2026

Evaluation results

MethodLinks
Qwen3 235BAccessibility=Open-Source, Parameters=235B2026.0510097.11
Qwen3 32BAccessibility=Open-Source, Parameters=32B2026.0510090.72
Qwen3 MaxAccessibility=Closed-Source2026.0510094.73
Qwen3 30BAccessibility=Open-Source, Parameters=30B2026.0599.9792.86
Qwen3 8BAccessibility=Open-Source, Parameters=8B2026.0599.9786.89
Deepseek V3.2Accessibility=Open-Source2026.0599.9792.86
Qwen3 14BAccessibility=Open-Source, Parameters=14B2026.0599.9587.59
Qwen3 Coder PlusAccessibility=Closed-Source2026.0599.9594.03
Qwen3 Next 80BAccessibility=Open-Source, Parameters=80B2026.0599.9294.79
GLM 4.7Accessibility=Open-Source2026.0599.9294.86
Gemini 3 ProAccessibility=Closed-Source2026.0599.6498.7
MiniMax M2.1Accessibility=Open-Source2026.0599.5695.02
GPT OSS 120BAccessibility=Open-Source, Parameters=120B2026.0599.4396.58
Seed 1.8Accessibility=Closed-Source, Version=1.82026.0599.497.45
Kimi Linear 48BAccessibility=Open-Source, Parameters=48B2026.0599.2248.46
Qwen3 4BAccessibility=Open-Source, Parameters=4B2026.0598.8569.94
Kimi K2 ThinkingAccessibility=Open-Source2026.0598.296.06
Mimo V2 FlashAccessibility=Open-Source2026.0597.8480.42
GLM 4.7 FlashAccessibility=Open-Source2026.0597.8156.8
Ministral 3 14BAccessibility=Open-Source, Parameters=14B2026.0597.1872.26
GPT OSS 20BAccessibility=Open-Source, Parameters=20B2026.0594.3490.56
Ministral 3 8BAccessibility=Open-Source, Parameters=8B2026.0592.2346.35
Seed OSS 36BAccessibility=Open-Source, Parameters=36B2026.0591.1687.96