List your company

Financial Numerical Reasoning on FinQA (test)

84.66Execution AccuracyDSA-7B
32.202445.821259.4473.0588May 29, 2026May 31, 2026Jun 2, 2026Jun 4, 2026Jun 6, 2026Jun 8, 2026Jun 10, 2026

Evaluation results

MethodLinks
DSA-7BModel Scale=7B2026.0584.66
DSA-3BModel Scale=3B2026.0579.12
MOCA-AGENTBackbone=Qwen3.6-27B, Type=Open-source LLMs2026.0678.29
FINDERBackbone=Mistral-7B Instruct v0.22026.0575.32
Fino1-14BType=Open-source LLMs2026.0674.18
DSA-1.5BModel Scale=1.5B2026.0573.5
Qwen2.5-72BType=Open-source LLMs2026.0673.38
DeepSeek-V3Type=Proprietary LLMs2026.0673.2
Qwen2.5-32BType=Open-source LLMs2026.0673.11
Fino1-7BType=Open-source LLMs2026.0673.03
GPT-4oType=Proprietary LLMs2026.0672.49
APOLLO2026.0571.07
ENCORE2026.0569.4
PoT-GPT-4Backbone=GPT-42026.0569.38
GPT-4.5Type=Proprietary LLMs2026.0668.94
Llama-3.3-70BType=Open-source LLMs2026.0668.15
CBR-Ren2026.0567.81
DS-R1-Distill-Llama-70BType=Open-source LLMs2026.0666.73
DeepSeek-R1Type=Proprietary LLMs2026.0665.13
ArgRecog2026.0564.86
Qwen3-32BType=Open-source LLMs2026.0664.15
FinQANet2026.0561.24
GPT-o3-miniType=Proprietary LLMs2026.0660.87
Qwen2.5-7B-InsZero-shot=true, Model Scale=7B2026.0560.8
GPT-o1-previewType=Proprietary LLMs2026.0649.07
Qwen2.5-3B-InsZero-shot=true, Model Scale=3B2026.0547.5
Qwen2.5-1.5B-InsZero-shot=true, Model Scale=1.5B2026.0534.22