Library · established

black-forest-labs/flux2

Minimal inference code for open-weight FLUX.2 image generation/editing, including Klein 4B/9B and larger dev variants.

image generationimage editinggenerative AI
Installation GuideInstruction Manual

What it contains

Minimal inference code for open-weight FLUX.2 image generation/editing, including Klein 4B/9B and larger dev variants.

Typical uses

Run local/self-managed text-to-image and reference-image editing; choose models based on speed, quality, tuning, hardware, and license.

Why people choose it

Inference: chosen when local/open-weight model control matters more than a hosted image API.

License
Mixed model licenses
Maintenance
active
Latest release
Unknown
Last meaningful update
Unknown
Maturity
established
Production readiness
production

Classification

Subject domains

image generationimage editinggenerative AI

Task categories

image generationimage editinggenerative AI

Project phases

implementationoperations

Secondary types

model repositoryCLI

Select when

  • Need local/open-weight image generation/editing

Do not select when

  • Model license incompatible
  • No suitable GPU
Implementation complexity
high
Setup effort
high
Learning curve
high
Confidence
high

Routing heuristic 80.0%

Routing specificity
4/5
Setup simplicity
1/5
Operational maturity
5/5
Composability
4/5
Local control
3/5
Security boundary clarity
5/5
First-party routing documentation
5/5

Strengths

None documented.

Poor-fit scenarios

  • Model license incompatible
  • No suitable GPU

Limitations

None documented.

Klein 4B library
README.md

Fast 4B generation/editing; Apache-2.0 model variants.

generationediting
Klein 9B library
README.md

Higher-capacity non-commercial variants.

generationediting
FLUX.2 dev library
README.md

32B high-quality research/tuning model.

generationeditingfine-tuning
scripts/cli.py

Interactive generation/editing.

CLI
docs/flux2_with_prompt_upsampling.md

Expand prompts locally/OpenRouter.

prompt enhancement
Languages
Python
Frameworks
PyTorch
Runtimes
Python 3.12
Operating systems
None documented
Installation methods
None documented
Interfaces
None documented
Required credentials
None documented
External services
Hugging Face, optional OpenRouter
Hardware
~8GB VRAM cited for Klein 4B, high VRAM for dev
Major dependencies
None documented

Compatibility notes

None documented.

One-sentence semantic summary

Minimal inference code for open-weight FLUX.2 image generation/editing, including Klein 4B/9B and larger dev variants.

Capability keywords

text-to-imagesingle/multi-reference editfine-tuningprompt upsampling

User intent phrases

  • Need local/open-weight image generation/editing

Negative match phrases

  • Model license incompatible
  • No suitable GPU

Differentiators

None documented.

Security considerations

  • Check model-specific license; consider provenance/watermark requirements.

Privacy considerations

None documented.

Uncertainty

Unresolved items

None documented.

Inference notes

None documented.

Evidence

Repository usage guidance

installation summary
Unknown
basic usage summary
Run local/self-managed text-to-image and reference-image editing; choose models based on speed, quality, tuning, hardware, and license.
documented entry points
None documented
key configuration files
None documented
important directories
scripts/, src/, docs/
documentation paths
README.md
example paths
None documented

Relationships

No explicit repository relationships indexed.

Recommended combinations

Not part of a curated combination.

Routing rules

rule_036: Need local/open-weight image generation/editing P964

Rationale

Minimal inference code for open-weight FLUX.2 image generation/editing, including Klein 4B/9B and larger dev variants.

Required conditions

None documented.

Preferred

black-forest-labs/flux2

Fallback

None