Codesota · Computer Code · Code Generation · MBPPTasks/Computer Code/Code Generation
Code Generation · benchmark dataset · 2021 · PYTHON

Mostly Basic Python Problems.

974 crowd-sourced Python programming problems suitable for beginners. Covers programming fundamentals and standard library.

Paper Download datasetSubmit a result
§ 01 · Leaderboard

Best published scores.

19 results indexed across 1 metric. Shaded row marks current SOTA; ties broken by submission date.


Primary
pass@1 · higher is better
pass@1· primary
19 rows
#ModelOrgSubmittedPaper / codepass@1
01o4-miniAPIOpenAIMar 2026official-model-card94.90
02o3-miniAPIOpenAIMar 2026official-model-card93.30
03Claude Opus 4APIAnthropicMar 2026official-model-card92
04GPT-4.1APIOpenAIMar 2026official-model-card90.90
05Qwen2.5-Coder 32BOSSAlibabaSep 2024Qwen2.5-Coder Technical Report · code90.20
06Claude Sonnet 4APIAnthropicMar 2026official-model-card89.60
07DeepSeek-Coder-V2-InstructOSSDeepSeekSep 2024arxiv-2409.1218689.40
08DeepSeek-Coder-V2-InstructOSSDeepSeekJun 2024DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source… · code89.40
09DeepSeek-V3OSSDeepSeekMar 2026arxiv89.30
10Claude 3.5 SonnetAPIAnthropicDec 2025anthropic-blog89.20
11GPT-4oAPIOpenAIDec 2025openai-blog87.80
12Llama-4-MaverickOSSMetaApr 2025meta-model-card77.60
13Codestral 22BMistralMay 2024official-blog75.40
14Gemma-3-27bGoogleMar 2025arxiv-2503.1978674.40
15Gemma 3 12B ITGoogle DeepMindMar 2025arxiv-2503.1978673
16Llama-4-ScoutOSSMetaApr 2025meta-model-card67.80
17Gemma 3 4B ITGoogle DeepMindMar 2025arxiv-2503.1978663.20
18Code Llama 34BOSSMetaMar 2026arxiv62.60
19StarCoder2 15BOSSBigCodeFeb 2024StarCoder2 and The Stack v2: The Next Generation · code54.40
Fig 2 · Rows sorted by score within each metric. Shaded row marks SOTA. Dates reflect model or paper release where available, otherwise the date Codesota accessed the source.
§ 03 · Progress

5 steps
of state of the art.

Each row below marks a model that broke the previous record on pass@1. Intermediate submissions are kept in the leaderboard above; only SOTA-setting entries are re-listed here.

Higher scores win. Each subsequent entry improved upon the previous best.

SOTA line · pass@1
  1. Feb 29, 2024StarCoder2 15BBigCode54.40
  2. May 29, 2024Codestral 22BMistral75.40
  3. Jun 17, 2024DeepSeek-Coder-V2-InstructDeepSeek89.40
  4. Sep 19, 2024Qwen2.5-Coder 32BAlibaba90.20
  5. Mar 27, 2026o4-miniOpenAI94.90
Fig 3 · SOTA-setting models only. 5 entries span Feb 2024 Mar 2026.
§ 04 · Literature

3 papers
tied to this benchmark.

Every paper below corresponds to at least one row in the leaderboard above. Click through for the arXiv preprint and, when available, the reference implementation.

§ 06 · Contribute

Have a score that beats
this table?

Submit a checkpoint and a reproduction script. We will run it, publish the score, and — if it takes the top — annotate the step on the progress chart with your name.

Submit a result Read submission guide
What a submission needs
  • 01A public checkpoint or API endpoint
  • 02A reproduction script with frozen commit + seed
  • 03Declared evaluation environment (Python, deps)
  • 04One row per metric declared by this dataset
  • 05A contact so we can follow up on discrepancies
MBPP — Code Generation benchmark · Codesota | CodeSOTA