MBPP
MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.
Progress Over Time
Interactive timeline showing model performance evolution on MBPP
State-of-the-art frontier
Open
Proprietary
MBPP Leaderboard
33 models
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 |  Sarvam AI | 30B | — | — | ||
| 2 | 50B | — | — | |||
| 3 |  Alibaba Cloud / Qwen Team | 32B | 128K | $0.09 / $0.09 | ||
| 4 |  OpenBMB | 9B | — | — | ||
| 5 | Alibaba Cloud / Qwen Team | 73B | 131K | $0.35 / $0.40 | ||
| 6 | 8B | — | — | |||
| 7 | Alibaba Cloud / Qwen Team | 33B | — | — | ||
| 7 | Alibaba Cloud / Qwen Team | 34B | — | — | ||
| 9 | Alibaba Cloud / Qwen Team | 7B | — | — | ||
| 10 | Alibaba Cloud / Qwen Team | 15B | — | — | ||
| 11 | Alibaba Cloud / Qwen Team | 235B | 128K | $0.10 / $0.10 | ||
| 12 |  Microsoft | 60B | — | — | ||
| 13 | Alibaba Cloud / Qwen Team | 72B | — | — | ||
| 14 | Alibaba Cloud / Qwen Team | 8B | 131K | $0.30 / $0.30 | ||
| 15 |  Mistral AI | 22B | — | — | ||
| 16 |  Meta | 400B | 1.0M | $0.17 / $0.60 | ||
| 17 |  Google | — | — | — | ||
| 18 | Mistral AI | 24B | — | — | ||
| 19 | Google | 27B | 131K | $0.10 / $0.20 | ||
| 20 | Alibaba Cloud / Qwen Team | 7B | — | — | ||
| 21 | Google | 12B | 131K | $0.05 / $0.10 | ||
| 22 | Mistral AI | 24B | — | — | ||
| 23 | Microsoft | 4B | 128K | $0.10 / $0.10 | ||
| 24 | Meta | 109B | 10.0M | $0.08 / $0.30 | ||
| 25 | Alibaba Cloud / Qwen Team | 8B | — | — | ||
| 26 | 2B | — | — | |||
| 26 | Google | 8B | 32K | $20.00 / $40.00 | ||
| 28 | Google | 4B | 131K | $0.02 / $0.04 | ||
| 29 | Google | 27B | — | — | ||
| 30 | 2B | — | — | |||
| 30 | Google | 8B | — | — | ||
| 32 | Google | 9B | — | — | ||
| 33 | Google | 1B | — | — |
Notice missing or incorrect data?
FAQ
Common questions about MBPP
MBPP (Mostly Basic Python Problems) is a benchmark of 974 crowd-sourced Python programming problems designed to be solvable by entry-level programmers. Each problem consists of a task description, code solution, and 3 automated test cases covering programming fundamentals and standard library functionality.
The MBPP paper is available at https://arxiv.org/abs/2108.07732. This paper provides detailed information about the benchmark methodology, dataset creation, and evaluation criteria.
The MBPP leaderboard ranks 33 AI models based on their performance on this benchmark. Currently, Sarvam-30B by Sarvam AI leads with a score of 0.927. The average score across all models is 0.741.
The highest MBPP score is 0.927, achieved by Sarvam-30B from Sarvam AI.
33 models have been evaluated on the MBPP benchmark, with 0 verified results and 33 self-reported results.
MBPP is categorized under general and reasoning. The benchmark evaluates text models.