Benchmark
GSM8k
Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary…
- Released
- 2021-04-01
- Modality
- text
- Categories
- math, reasoning
- Openness
- unknown
- Source
- llm_stats
- Reported scores
- 48
Reported scores
llm_stats
| Model | Organization | Reported value | Reported | Evidence |
|---|---|---|---|---|
| Claude 3 Haiku | Anthropic | 0.889 | 2024-03-13 | evidence |
| Claude 3 Opus | Anthropic | 0.95 | 2024-02-29 | evidence |
| Claude 3 Sonnet | Anthropic | 0.923 | 2024-02-29 | evidence |
| Claude 3.5 Sonnet | Anthropic | 0.964 | 2024-06-21 | evidence |
| Claude 3.5 Sonnet | Anthropic | 0.964 | 2024-10-22 | evidence |
| Command R+ | Cohere | 0.707 | 2024-08-30 | evidence |
| DeepSeek-V2.5 | DeepSeek | 0.951 | 2024-05-08 | evidence |
| ERNIE 4.5 | Baidu | 0.252 | 2025-06-25 | evidence |
| Gemini 1.5 Flash | 0.862 | 2024-05-01 | evidence | |
| Gemini 1.5 Pro | 0.908 | 2024-05-01 | evidence | |
| Gemma 2 27B | 0.74 | 2024-06-27 | evidence | |
| Gemma 2 9B | 0.686 | 2024-06-27 | evidence | |
| Gemma 3 12B | 0.944 | 2025-03-12 | evidence | |
| Gemma 3 1B | 0.628 | 2025-03-12 | evidence | |
| Gemma 3 27B | 0.959 | 2025-03-12 | evidence | |
| Gemma 3 4B | 0.892 | 2025-03-12 | evidence | |
| GPT-4.5 | OpenAI | 0.97 | 2025-02-27 | evidence |
| Granite 3.3 8B Base | IBM | 0.59 | 2025-04-16 | evidence |
| Granite 3.3 8B Instruct | IBM | 0.8089 | 2025-04-16 | evidence |
| Grok-1.5 | xAI | 0.9 | 2024-03-28 | evidence |
| IBM Granite 4.0 Tiny Preview | IBM | 0.701 | 2025-05-02 | evidence |
| Jamba 1.5 Large | AI21 Labs | 0.87 | 2024-08-22 | evidence |
| Jamba 1.5 Mini | AI21 Labs | 0.758 | 2024-08-22 | evidence |
| Kimi K2 Base | Moonshot AI | 0.921 | 2025-07-11 | evidence |
| Kimi K2 Instruct | Moonshot AI | 0.973 | 2025-07-11 | evidence |
| Llama 3.1 405B Instruct | Meta | 0.968 | 2024-07-23 | evidence |
| Llama 3.1 Nemotron 70B Instruct | NVIDIA | 0.9143 | 2024-10-01 | evidence |
| Llama 3.2 3B Instruct | Meta | 0.777 | 2024-09-25 | evidence |
| MiMo-V2.5-Pro | Xiaomi | 0.996 | 2026-04-27 | evidence |
| Mistral Large 2 | Mistral | 0.93 | 2024-07-24 | evidence |
| Mistral Small 3 24B Base | Mistral | 0.8073 | 2025-01-30 | evidence |
| Nova Lite | Amazon | 0.945 | 2024-11-20 | evidence |
| Nova Micro | Amazon | 0.923 | 2024-11-20 | evidence |
| Nova Pro | Amazon | 0.948 | 2024-11-20 | evidence |
| o1 | OpenAI | 0.971 | 2024-12-17 | evidence |
| Phi 4 Mini | Microsoft | 0.886 | 2025-02-01 | evidence |
| Phi-3.5-mini-instruct | Microsoft | 0.862 | 2024-08-23 | evidence |
| Phi-3.5-MoE-instruct | Microsoft | 0.887 | 2024-08-23 | evidence |
| Qwen2 72B Instruct | Qwen | 0.911 | 2024-07-23 | evidence |
| Qwen2 7B Instruct | Qwen | 0.823 | 2024-07-23 | evidence |
| Qwen2.5 14B Instruct | Qwen | 0.948 | 2024-09-19 | evidence |
| Qwen2.5 32B Instruct | Qwen | 0.959 | 2024-09-19 | evidence |
| Qwen2.5 72B Instruct | Qwen | 0.958 | 2024-09-19 | evidence |
| Qwen2.5 7B Instruct | Qwen | 0.916 | 2024-09-19 | evidence |
| Qwen2.5-Coder 32B Instruct | Qwen | 0.911 | 2024-09-19 | evidence |
| Qwen2.5-Coder 7B Instruct | Qwen | 0.839 | 2024-09-19 | evidence |
| Qwen2.5-Omni-7B | Qwen | 0.887 | 2025-03-27 | evidence |
| Qwen3 235B A22B | Qwen | 0.9439 | 2025-04-29 | evidence |
Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.