# facebookresearch/chameleon

Repository for Meta Chameleon, a mixed-modal early-fusion foundation model from FAIR.

Repository: https://github.com/facebookresearch/chameleon
Canonical: https://ross.abutalabs.com/products/facebookresearch-chameleon
Homepage: https://arxiv.org/abs/2405.09818
Language: Python
License: NOASSERTION
License Family: other
Archived: true
Last push: 2024-07-29T22:36:44+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 57
- inputs: {"age_days": 807, "days_push": 765, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2103, forks 118 (observed 2026-08-28T04:06:13.732777+00:00)

## What it is
Repository for Meta Chameleon, an early-fusion token-based mixed-modal foundation model that understands and generates interleaved images and text. It provides GPU inference code, a browser-based multimodal input/output viewer, a lightweight miniviewer, and evaluation prompts for the released 7B and 30B checkpoints.

## Use cases
- run inference with a mixed-modal image-and-text foundation model
- generate images and text interleaved from a single model
- visualize multimodal model inputs and outputs in a browser
- evaluate a multimodal model on visual question answering and captioning
- experiment with early-fusion token-based multimodal architectures
- download and serve Meta Chameleon checkpoints locally

## When to choose
- you need a single model that both understands and generates images and text
- you want to reproduce or study the Chameleon paper's early-fusion approach
- you have CUDA GPUs and want fast local inference of the 7B or 30B checkpoints
- you need a tool to inspect mixed-modal prompts and generations

## When to avoid
- you only need text-only LLM inference with broad ecosystem support
- you lack a CUDA-capable GPU and cannot use the HuggingFace implementations
- you need a production-ready, actively maintained multimodal API
- you cannot accept the custom research license or checkpoint access requirements

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-inference, machine-learning, image-processing, nlp
- domain: large-language-models, artificial-intelligence, deep-learning, machine-learning
- platform: python
- tags: multimodal, foundation-model, early-fusion, image-generation, text-generation, meta-ai, research-model, inference, gpu, docker, linux

## Member repositories
- facebookresearch/chameleon (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:13.732777+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 2026-08-30T02:54:14.376376+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/chameleon (fetched 2026-08-28T04:06:13.732777+00:00, sha 0dcfda8d7779)
  - homepage: https://arxiv.org/abs/2405.09818 (fetched 2026-08-29T10:34:29.202232+00:00, sha a0d13976ec45)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T10:34:29.244218+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T10:34:29.248471+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T10:34:29.250312+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T10:34:29.246627+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.