# jina-ai/dalle-flow

🌊 A Human-in-the-Loop workflow for creating HD images from text

Repository: https://github.com/jina-ai/dalle-flow
Canonical: https://ross.abutalabs.com/products/dalle-flow
Language: Python
License Family: other
Topics: dalle, generative-art, human-in-the-loop, jina, neural-search, openai, swinir, dalle-mega, dalle-mini, glid3
Last push: 2023-05-16T09:54:02+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1586, "days_push": 1205, "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 2829, forks 206 (observed 2026-08-28T04:07:24.270068+00:00)

## What it is
DALL·E Flow is a client-server workflow for generating high-definition images from text prompts using DALL·E-Mega, GLID-3 XL, and Stable Diffusion, with CLIP-based candidate ranking, diffusion refinement, and SwinIR/RealESRGAN upscaling. It formalizes generative art as an iterative human-in-the-loop process built on the Jina framework with gRPC/Websocket/HTTP access.

## Use cases
- generate HD images from text prompts
- iteratively refine AI-generated art with human feedback
- rank text-to-image candidates with CLIP
- upscale generated images to 1024x1024
- self-host a text-to-image generation server
- compare DALL-E Mega, GLID-3 XL, and Stable Diffusion outputs

## When to choose
- you want an interactive, multi-step text-to-image pipeline rather than single-prompt output
- you need a self-hosted image generation server with gRPC/HTTP/Websocket APIs
- you want automatic candidate ranking, diffusion refinement, and upscaling in one workflow

## When to avoid
- you need a maintained or supported project - it is deprecated and has no license
- you want a simple one-shot text-to-image tool without server deployment
- you lack GPU resources, since the pipeline requires heavy model inference

## Facets
- artifact type: service
- maturity: maintenance
- function: machine-learning, image-processing, workflow-automation, http-server, rpc
- domain: artificial-intelligence, image-processing, deep-learning
- platform: python, self-hosted
- tags: text-to-image, dalle, stable-diffusion, human-in-the-loop, generative-art, image-upscaling, clip, jina, client-server, docker, gpu, web-server

## Member repositories
- jina-ai/dalle-flow (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:24.270068+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 2026-08-30T07:38:11.558453+00:00, confidence not recorded.
  - readme: https://github.com/jina-ai/dalle-flow (fetched 2026-08-28T04:07:24.270068+00:00, sha 11eeeeeb3655)
- Data as of 2026-08-30T08:39:29.467469+00:00.