# elixir-nx/bumblebee

Pre-trained Neural Network models in Axon (+ 🤗 Models integration)

Repository: https://github.com/elixir-nx/bumblebee
Canonical: https://ross.abutalabs.com/products/elixir-nx-bumblebee
Language: Elixir
License: Apache-2.0
License Family: permissive
Topics: axon, elixir, nx, hugging-face, pre-trained, machine-learning, transformer
Last push: 2026-08-19T02:21:19+00:00

## Health v2 (maintenance only)
Score: 89/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 98, release rhythm 70, longevity 100
- inputs: {"age_days": 1583, "days_push": 15, "days_rel": 42, "gap_med": 157.5, "n_releases_24m": 5}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1662, forks 145 (observed 2026-08-28T04:05:18.731480+00:00)

## What it is
Bumblebee is an Elixir library providing pre-trained neural network models built on Axon, with integration for downloading models from Hugging Face. It offers high-level serving APIs to run machine learning tasks like text classification, generation, image classification, and speech recognition with a few lines of code.

## Use cases
- run hugging face transformer models in elixir
- serve bert for text classification in a phoenix app
- generate text with gpt2 in elixir
- run stable diffusion image generation in elixir
- do speech-to-text with whisper in elixir
- embed text for semantic search in elixir
- deploy machine learning models in livebook notebooks

## When to choose
- you are building in Elixir and want pre-trained models without leaving the ecosystem
- you want to integrate ML serving into Phoenix/LiveView apps
- you want GPU-accelerated inference via EXLA/XLA
- you want to experiment with models interactively in Livebook

## When to avoid
- you need the full PyTorch/TensorFlow training ecosystem
- you need model architectures not yet supported by Bumblebee
- you are not working in the Elixir/BEAM ecosystem
- you need fine-grained low-level control over model internals

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, llm-inference, nlp, image-processing, speech-recognition, stable-diffusion, transformers
- domain: machine-learning, deep-learning, large-language-models, artificial-intelligence
- platform: elixir, windows
- tags: hugging-face, axon, nx, pre-trained-models, exla, gpu, serving, elixir-nx, natural-language-processing, beam, linux, macos

## Member repositories
- elixir-nx/bumblebee (main) score 89

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:18.731480+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 2026-08-30T03:44:33.987164+00:00, confidence not recorded.
  - readme: https://github.com/elixir-nx/bumblebee (fetched 2026-08-28T04:05:18.731480+00:00, sha 6b567a4a2bd0)
- Data as of 2026-08-30T08:39:29.467469+00:00.