# declare-lab/tango

A family of diffusion models for text-to-audio generation.

Repository: https://github.com/declare-lab/tango
Canonical: https://ross.abutalabs.com/products/declare-lab-tango
Homepage: https://tango2-web.github.io/
Language: Python
License: NOASSERTION
License Family: other
Topics: audio-generation, diffusion, diffusion-models, language-models, large-language-models, text-to-audio
Last push: 2025-07-29T02:13:29+00:00

## Health v2 (maintenance only)
Score: 45/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 34, release rhythm 35, longevity 88
- inputs: {"age_days": 1241, "days_push": 401, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1239, forks 105 (observed 2026-08-28T04:04:06.033258+00:00)

## What it is
Tango is a family of latent diffusion models for text-to-audio generation, with Tango 2 improving prompt alignment via DPO-based fine-tuning. It provides Python code and pretrained checkpoints on Hugging Face for generating sound effects and audio from text prompts.

## Use cases
- generate sound effects from text descriptions
- create audio for film or game pre-production mockups
- text-to-audio generation with diffusion models
- fine-tune a text-to-audio model with preference optimization
- generate audio with temporally ordered events from a prompt

## When to choose
- you need research-grade text-to-audio generation in Python
- you want pretrained checkpoints and DPO-aligned audio quality
- you need reproducible models from published papers

## When to avoid
- you need fast production inference - TangoFlux is the faster SOTA successor
- you need lyrics-to-song generation - use the Jam project instead
- you need a hosted no-code tool rather than a Python library

## Facets
- artifact type: library
- maturity: active
- function: audio-processing, machine-learning, deep-learning, llm-training
- domain: deep-learning, artificial-intelligence, media
- platform: python, cross-platform
- tags: text-to-audio, diffusion-models, latent-diffusion, dpo, audio-generation, huggingface, audio, gpu

## Member repositories
- declare-lab/tango (main) score 45

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:06.033258+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 2026-08-30T05:09:57.107533+00:00, confidence not recorded.
  - readme: https://github.com/declare-lab/tango (fetched 2026-08-28T04:04:06.033258+00:00, sha bf4a8bac906c)
  - homepage: https://tango2-web.github.io/ (fetched 2026-08-29T12:20:56.035198+00:00, sha 457ec3e7a512)
- Data as of 2026-08-30T08:39:29.467469+00:00.