# BabitMF/bmf

Cross-platform, customizable multimedia/video processing framework.  With strong GPU acceleration, heterogeneous design, multi-language support, easy to use, multi-framework compatible and high performance, the framework is ideal for transcoding, AI inference, algorithm integration, live video streaming, and more.

Repository: https://github.com/BabitMF/bmf
Canonical: https://ross.abutalabs.com/products/bmf
Homepage: https://babitmf.github.io/
Language: C++
License: Apache-2.0
License Family: permissive
Topics: bmf, bytedance, cpp, cross-platform, python, ai, arm, cuda, gpu, heterogeneous, mediacodec, multimedia, nvidia, tensorrt, transcode, x86-64, ffmpeg, live-video, numpy, opencv
Last push: 2026-08-26T04:49:22+00:00

## Health v2 (maintenance only)
Score: 75/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 40, longevity 81
- inputs: {"age_days": 1145, "days_push": 7, "days_rel": 432, "gap_med": 18, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1034, forks 107 (observed 2026-08-28T04:03:18.601780+00:00)

## What it is
BMF (Babit Multimedia Framework) is a cross-platform, multi-language multimedia and video processing framework developed by ByteDance, offering Python, C++, and Go APIs with strong GPU acceleration and heterogeneous computing support. It is used in production for transcoding, live video, cloud editing, and AI inference, processing billions of videos daily.

## Use cases
- transcode video files with gpu acceleration
- build a video processing pipeline in python
- integrate ai inference into video workflows
- process live video streams
- convert between ffmpeg numpy pytorch opencv tensor formats
- develop custom multimedia processing modules
- run video preprocessing on mobile devices

## When to choose
- you need high-performance transcoding or live video processing with gpu acceleration
- you want a multi-language (python/c++/go) multimedia framework with modular extensibility
- you need to combine video processing with ai inference in one pipeline
- you need cross-platform support including x86, arm, and mobile

## When to avoid
- you only need simple one-off ffmpeg commands without a framework
- your project is limited to a single language or platform unsupported by bmf
- you need a lightweight pure-python-only video library

## Facets
- artifact type: framework
- maturity: active
- function: video-processing, audio-processing, image-processing, machine-learning, gpu-computing, streaming, sdk
- domain: media, machine-learning, cross-platform, developer-tools
- platform: windows, cross-platform, python, cpp
- tags: multimedia, transcoding, ffmpeg, tensorrt, cuda, mediacodec, opencv, pytorch, pipeline, bytedance, bmf-lite, video, linux, macos, android, gpu

## Member repositories
- BabitMF/bmf (main) score 75

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:18.601780+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 2026-08-30T07:06:29.874728+00:00, confidence not recorded.
  - readme: https://github.com/BabitMF/bmf (fetched 2026-08-28T04:03:18.601780+00:00, sha 946b404bc5fc)
  - homepage: https://babitmf.github.io/ (fetched 2026-08-29T13:05:47.460908+00:00, sha a2a848efe4b6)
- Data as of 2026-08-30T08:39:29.467469+00:00.