List your company

Mathematical Reasoning on MGSM Rev2

99.4Overall Result AccuracyGEMINI-3-FLASH
9.54432.87256.279.528Apr 6, 2026Apr 18, 2026May 1, 2026May 13, 2026May 26, 2026Jun 7, 2026Jun 20, 2026

Evaluation results

MethodLinks
GEMINI-3-FLASHResource Profile=Low2026.0699.4———————0.599.5
GEMINI-3-FLASHResource Profile=Standard2026.0699.3———————0.499.6
CLAUDE-SONNET-4.6Model Version=4.62026.0698.7———————-2.699.4
GEMMA-4-31B-ITModel Variant=31B-IT2026.0698.3———————1.498.4
GEMINI-2.5-FLASHThinking=On2026.0698.2———————1.398.7
GPT-OSS-120BParameter Count=120B2026.0697———————2.697.4
GEMMA-4-26B-A4B-ITModel Variant=26B-A4B-IT2026.0696.5———————396.8
GPT-OSS-20BParameter Count=20B2026.0695.2———————3.896.1
GEMMA-3-27BParameter Count=27B2026.0693.3———————4.195.3
CLAUDE-OPUS-4.7Model Version=4.72026.0690.4———————-197.3
QWEN-3-30B-A3BModel Variant=30B-A3B2026.0688.6———————9.889.8
GEMMA-3-12BParameter Count=12B2026.0688.1———————7.991.3
GEMINI-3-FLASHResource Profile=Minimal2026.0686.5———————0.496.2
QWEN-3-4BParameter Count=4B2026.0679.4———————18.181.3
QWEN3 4Bn=42026.0475.574.775.575.573.475.675.973.7——
QWEN3 4Bn=82026.0475.474.875.675.673.574.674.673.2——
QWEN3 4Bn=162026.0475.47475.575.873.274.874.773——
QWEN3 8Bn=322026.0474.773.875.675.172.675.674.171.7——
QWEN3 4Bn=322026.0474.674.575.375.372.376.174.372.7——
QWEN3 8Bn=162026.0474.473.874.174.373.374.973.972.1——
QWEN3 8Bn=82026.0474.173.274.37473.87473.872.8——
QWEN3 8Bn=42026.0473.973.373.974.673.574.173.672.1——
GEMMA-3-4BParameter Count=4B2026.0672.9———————13.482.6
DISTILL-QWEN 8Bn=162026.0462.662.561.563.263.263.96060.7——
DISTILL-QWEN 8Bn=42026.0462.362.762.36362.263.360.463.3——
DISTILL-QWEN 8Bn=322026.0461.562.358.96461.765.160.959.6——
DISTILL-QWEN 8Bn=82026.0461.262.362.262.662.964.161.263.6——
GEMINI-2.5-FLASHThinking=Off2026.0655.6———————2.889.8
CLAUDE-HAIKU-4.5Model Version=4.52026.0654———————-1.386
DISTILL-QWEN 1.5Bn=322026.0436.836.937.1433640.239.134.3——
DISTILL-QWEN 1.5Bn=162026.0436.136.536.540.236.139.638.735.1——
DISTILL-QWEN 1.5Bn=42026.043636.636.737.836.738.138.136——
DISTILL-QWEN 1.5Bn=82026.043636.836.739.536.738.738.334.9——
GEMMA-3-1BParameter Count=1B2026.0625.9———————24.747.7
GEMMA-4-E4B-ITModel Variant=E4B-IT2026.0617.5———————4.468
GEMMA-4-E2B-ITModel Variant=E2B-IT2026.0613———————553.7