GSM8K Leaderboard

Grade School Math 8K — 8,500 high quality grade school math word problems.

As of April 18, 2026, the top-scoring model on GSM8K is Claude Opus 4 at 96.2%, followed by Claude Opus 4 at 96.2% and R1 at 96.2%. 56 models have been evaluated on this benchmark.

Last updated: April 18, 2026

Models

Best Score

96.2

Average

91.2

Std Dev

10.0

Categories

Mathematical Problem Solving

SourceLayerLens

Provider	Model	Input $/M	Output $/M	GSM8K	Actions
A Anthropic	Claude Opus 4 Thinking	$15.000	$75.000	96.2	Try
A Anthropic	Claude Opus 4	$15.000	$75.000	96.2	Try
DS DeepSeek	R1	$0.550	$2.000	96.2	Try
O OpenAI	o4 Mini High	$1.100	$4.400	96.0	Try
A Anthropic	Claude 3.7 Sonnet	$3.000	$15.000	95.8	Try
A Anthropic	Claude Sonnet 4 Thinking	$3.000	$15.000	95.7	Try
A Anthropic	Claude Sonnet 4	$3.000	$15.000	95.7	Try
G Google	Gemini 2.5 Pro	$1.000	$10.000	95.7	Try
NV NVIDIA	Nemotron 3 Nano 30B A3B Thinking	$0.050	$0.200	95.4	Try
NV NVIDIA	Nemotron 3 Nano 30B A3B	$0.050	$0.200	95.4	Try
A Anthropic	Claude 3.7 Sonnet Thinking	$3.000	$15.000	95.4	Try
DS DeepSeek	R1 0528	$0.500	$2.150	95.3	Try
O OpenAI	o3 Mini	$0.550	$2.200	95.2	Try
M Meta	Llama 4 Maverick	$0.150	$0.600	95.2	Try
AL Alibaba	Qwen3 235B A22B Thinking	$0.455	$0.900	95.2	Try
AL Alibaba	Qwen3 235B A22B	$0.455	$0.900	95.2	Try
G Google	Gemini 2.0 Flash	$0.100	$0.400	95.1	Try
K Kimi	Kimi K2 0711	$0.550	$2.200	95.0	Try
K Kimi	Kimi K2 0711	$0.550	$2.200	95.0	Try
O OpenAI	GPT-5	$1.250	$10.000	94.8	Try
O OpenAI	GPT-5	$1.250	$10.000	94.8	Try
O OpenAI	GPT-5	$1.250	$10.000	94.8	Try
O OpenAI	GPT-5	$1.250	$10.000	94.8	Try
P Prime Intellect	INTELLECT-3	$0.200	$1.100	94.8	Try
BD Baidu	ERNIE 4.5 300B A47B	$0.280	$0.900	94.6	Try
CO Cohere	Command A	$2.500	$10.000	94.6	Try
MM MiniMax	MiniMax M2.1	$0.290	$0.950	94.6	Try
AL Alibaba	Qwen3 32B Thinking	$0.080	$0.240	94.6	Try
AL Alibaba	Qwen3 32B	$0.080	$0.240	94.6	Try
MI Mistral	Mistral Medium 3	$0.400	$2.000	94.2	Try
AL Alibaba	Qwen3 Coder 30B A3B Instruct	$0.070	$0.270	94.2	Try
AL Alibaba	Qwen3 30B A3B Thinking	$0.080	$0.280	94.1	Try
AL Alibaba	Qwen3 30B A3B	$0.080	$0.280	94.1	Try
AL Alibaba	QwQ 32B	$0.150	$0.580	94.0	Try
AL Alibaba	QwQ 32B	$0.150	$0.580	94.0	Try
DS Deepseek	DeepSeek V3	$0.014	$0.028	93.8	Try
MM MiniMax	MiniMax M1	$0.400	$1.760	93.8	Try
MM MiniMax	MiniMax M1	$0.400	$1.760	93.8	Try
A Anthropic	Claude 3.5 Haiku	$0.800	$4.000	93.3	Try
X Xai	Grok 3 Beta	$3.000	$15.000	93.1	Try
AL Alibaba	Qwen2.5 72B Instruct	$0.120	$0.390	92.9	Try
MI Mistral	Pixtral Large 2411	$2.000	$6.000	91.9	Try
AM Amazon	Nova Pro 1.0	$0.800	$3.200	91.7	Try
MS Microsoft Azure	Phi 4	$0.065	$0.140	91.3	Try
AM Amazon	Nova Micro 1.0	$0.035	$0.140	91.1	Try
M Meta	Llama 3.1 405B Instruct	$0.900	$0.900	90.6	Try
MI Mistral	Devstral Small 1.1	$0.070	$0.280	89.2	Try
G Google	Gemma 3n 4B	$0.060	$0.120	88.9	Try
AL Alibaba	Qwen3 235B A22B Instruct 2507	$0.071	$0.100	88.7	Try
G Google	Gemini 2.5 Pro Preview 05-06	$1.250	$10.000	87.0	Try
G Google	Gemini 2.5 Flash Thinking	$0.300	$2.500	79.1	Try
G Google	Gemini 2.5 Flash Thinking	$0.300	$2.500	79.1	Try
G Google	Gemini 2.5 Flash	$0.300	$2.500	79.1	Try
G Google	Gemini 2.5 Flash	$0.300	$2.500	79.1	Try
IF Inflection	Inflection 3 Pi	$2.500	$10.000	64.0	Try
M Meta	Llama 3.2 3B Instruct	$0.030	$0.050	30.8	Try

Pricing from OpenRouter. Benchmarks from Artificial Analysis.

Get our weekly newsletter on pricing changes, new releases, and tools.

Join the Price Per Token Community

8 Ways to Use Fewer Tokens

About GSM8K

Grade School Math 8K — 8,500 high quality grade school math word problems.

This leaderboard shows all models with GSM8K benchmark scores, ranked from highest to lowest. Pricing data is included to help you compare performance against cost.