Share your thoughts, 1 month free Claude Pro on usSee more

Math on MATH-500 (Accuracy)

99.2Accuracy

GLM-4.7-Flash-T

Updated 12d ago

Evaluation Results

Method
GLM-4.7-Flash-T 2026.04	99.2	-
LongCat-Flash Exp-Chat 2025.12	98.8	-
GLM 4.6 2025.12	98.6	-
JoyAI-LLM Flash 2026.04	98.2	-
Qwen3-30B-A3B 2026.04	97.6	-
Latent-GRPO 2026.01	97.5	3,108.44
Qwen3-Next-80B-A3B 2026.04	97.4	-
DeepSeek V3.2 2025.12	97.2	-
Base 2026.01	97	-
GRPO (LLM-Judge) 2026.01	96.4	6,753.47
LongCat-Flash Chat 2025.12	96.4	-
Qwen3.5-35B-A3B 2026.04	95.8	-
Qwen3-8B 2026.03	94.8	-
SSA-LLM-8B 2026.03	94.8	-
Base 2026.01	93.4	-
Reasoning Memory 2026.04	92.6	-
Reasoning Memory 2026.04	92.3	-
Document RAG (Google) 2026.04	91.9	-
Latent-GRPO 2026.01	91.2	2,340.05
Reasoning Memory 2026.04	91.1	-
Reasoning Memory 2026.04	90.8	-
Length Scaling 2026.04	90.8	-
GRPO (LLM-Judge) 2026.01	90.6	4,988.84
No RAG 2026.04	90.5	-
Template RAG (Self-Discover) 2026.04	90.3	-
Length Scaling 2026.04	90.3	-
Template RAG (ReasonFlux) 2026.04	90.1	-
Document RAG (CompactDS) 2026.04	89.9	-
Trajectory RAG (Prefix) 2026.04	89.5	-
Trajectory RAG (Summary) 2026.04	89.5	-
Length Scaling 2026.04	87.6	-
Length Scaling 2026.04	87.3	-
AdaRAS 2026.01	86.4	-
Document RAG (Google) 2026.04	86.4	-
Trajectory RAG (Summary) 2026.04	85.5	-
OpenReasoning-Nemotron-1.5B 2026.01	85.4	-
Probing 2026.01	85.4	-
SSA-LLM-8B 2026.03	85.4	-
CoT 2026.01	84.8	-
Qwen3-8B 2026.03	84.6	-
Document RAG (CompactDS) 2026.04	84.6	-
No RAG 2026.04	84.1	-
Document RAG (Google) 2026.04	83.8	-
Reasoning Memory 2026.04	83.8	-
Reasoning Memory 2026.04	83.6	-
Template RAG (ReasonFlux) 2026.04	83	-
Template RAG (ReasonFlux) 2026.04	82.8	-
Trajectory RAG (Prefix) 2026.04	82.8	-
Template RAG (Self-Discover) 2026.04	82.5	-
Trajectory RAG (Summary) 2026.04	82.2	-
Template RAG (Self-Discover) 2026.04	82.1	-
OpenThinker-3-1.5B 2026.01	82	-
Kimi-Linear-48B-A3B-Instruct 2025.10	81.8	-
Document RAG (CompactDS) 2026.04	81.8	-
Length Scaling 2026.04	81.5	-
Kimi-Linear-48B-A3B-Instruct 2025.10	80.8	-
No RAG 2026.04	80.6	-
Length Scaling 2026.04	80.2	-
Trajectory RAG (Prefix) 2026.04	78.4	-
Base 2026.01	77.6	-
JoyAI-LLM Flash-Base 2026.04	77	-
Qwen3-1.7B-ALLMEM 2026.02	74.4	-
Qwen3-8B 2026.01	73.95	-
Qwen3-1.7B 2026.02	73.6	-
AutoMR 2025.10	69.6	-
Latent-GRPO 2026.01	67.3	2,280.52
rStar 2025.10	67	-
DeepSeek-R1-Distill-Qwen-1.5B 2026.01	66.2	-
Typhoon-S-8B 2026.01	65.93	-
GRPO (LLM-Judge) 2026.01	65.9	4,082.96
Meta-Reasoner 2025.10	65.4	-
MRP 2025.10	63.8	-
MaAS 2025.10	63.6	-
CoT 2025.10	61.6	-
Qwen3-30B-A3B-Base 2026.04	58	-
Qwen3.5-35B-A3B-Base 2026.04	54.8	-
AutoMR 2025.10	50.2	-
Qwen3-0.6B-ALLMEM 2026.02	49.8	-
Qwen3-0.6B 2026.02	48.8	-
rStar 2025.10	46.6	-
MaAS 2025.10	46.2	-
Meta-Reasoner 2025.10	44.4	-
MRP 2025.10	40.8	-
CoT 2025.10	36.8	-
Direct-I/O 2025.10	16.8	-
Direct-I/O 2025.10	12.6	-