# salesforce/CodeT5

Home of CodeT5: Open Code LLMs for Code Understanding and Generation

Repository: https://github.com/salesforce/CodeT5
Canonical: https://ross.abutalabs.com/products/codet5
Homepage: https://arxiv.org/abs/2305.07922
Language: Python
License: BSD-3-Clause
License Family: permissive
Topics: code-intelligence, code-generation, code-understanding, large-language-models, language-model
Archived: true
Last push: 2026-06-25T16:27:18+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 89, release rhythm 35, longevity 100
- inputs: {"age_days": 1843, "days_push": 69, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3093, forks 487 (observed 2026-08-28T04:07:42.405564+00:00)

## What it is
Official research release of CodeT5 and CodeT5+ open code large language models from Salesforce Research for code understanding and generation. It provides pretrained and fine-tuned checkpoints, training/evaluation code, and supports tasks like text-to-code generation, code autocompletion, and code summarization.

## Use cases
- generate code from natural language descriptions
- autocomplete function bodies given a function name
- summarize code functions in plain English
- fine-tune a code LLM on my own dataset
- download pretrained code language model checkpoints
- research on code intelligence and program synthesis

## When to choose
- you need open pretrained models for code generation or understanding tasks
- you want to fine-tune or benchmark code LLMs in research
- you need multilingual code summarization or completion models

## When to avoid
- you want a production-ready commercial coding assistant with IDE integration out of the box
- you need the newest general-purpose code LLM rather than these specific research models
- you have no GPU or ML infrastructure to run large models

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, llm-training, nlp, sdk
- domain: large-language-models, artificial-intelligence, developer-tools
- platform: python, cross-platform
- tags: code-generation, code-understanding, pretrained-models, encoder-decoder, huggingface, code-summarization, research-models, natural-language-processing, gpu

## Member repositories
- salesforce/CodeT5 (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:42.405564+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:27:19.682610+00:00, confidence not recorded.
  - readme: https://github.com/salesforce/CodeT5 (fetched 2026-08-28T04:07:42.405564+00:00, sha 02f64eaffa67)
  - homepage: https://arxiv.org/abs/2305.07922 (fetched 2026-08-29T09:42:35.621736+00:00, sha 2a281e44344f)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T09:42:35.624380+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T09:42:35.627662+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T09:42:35.629506+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T09:42:35.626091+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.