Skip to content

Black Forest Labs · Image

FLUX.2 [dev]

Black Forest Labs 🇩🇪ImageFLUX.1 family
From Cloudflare Workers AI's listing:
FLUX.2 [dev] is an image model from Black Forest Labs where you can generate highly realistic and detailed images, with multi-reference support.
llminfo IQ
92.4
Market median price
$0.024 /image
Listings
13
Providers
9

Evidence and sources

🤗 black-forest-labs/FLUX.2-dev Model architecture, as its sources state it llminfo IQ 92.4, ranked among image generation models. From 183 benchmarks. A fair amount of data behind it.How llminfo IQ is computedReleased 22 Nov 2025Maker HQ: DE (source)
Sources: LMArena data, Cloudflare Workers AI, Hugging FacePrimary data from LMArena from Cloudflare Workers AI: architecture, description from Hugging Face: release date

Advantages and trade-offs

Compared with 10 image generation models, picked mainly by popularity and IQ

Trade-offs

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 2 maker's prices; 9 market median prices.

Input & output

as stated by the model's providers
Text
Images
Images

Listings

 Cometapiblack-forest-labs/flux-2-devSources: CometAPI data, CometAPI Pricing dataNot published by the provider—$0.075
 NanoGPTflux-2-devSource: NanoGPTNot published by the provider$0.0204—
 NanoGPTflux-2-dev-image-to-image · image-to-imageSource: NanoGPTNot published by the provider$0.024—
 NanoGPTflux-2-dev-lora · loraSource: NanoGPTNot published by the provider$0.0306—
 NanoGPTflux-2-dev-lora-image-to-image · image-to-image, loraSource: NanoGPTNot published by the provider$0.051—
 Together AIblack-forest-labs/FLUX.2-devSources: LiteLLM data, Together AI dataNot published by the provider$0.0154—

7 listings without a price are hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Image generation

Arena ELO Text to ImageSource: LMArena data · Nov 2026
GEditBench v2Source: GEditBench v2
GenExam22 related benchmarks · 22 results
UniGenBench++136 related benchmarks · 136 results
Source: UniGenBench++ data
BenchmarkScore
(Chinese) - 2D
(Chinese Long) - 2D
(Chinese Long) - 3D
(Chinese Long) - Action-Overall
Show 132 more benchmarksHide 132 benchmarks
BenchmarkScore
(Chinese Long) - Animal
(Chinese Long) - Attribute-Overall
(Chinese Long) - Color
(Chinese Long) - Comparison
(Chinese Long) - Composition
(Chinese Long) - Compound-Overall
(Chinese Long) - Consistency
(Chinese Long) - Contact
(Chinese Long) - Expression
(Chinese Long) - Feature matching
(Chinese Long) - Full body
(Chinese Long) - Grammar-Overall
(Chinese Long) - Hand
(Chinese Long) - Imagination
(Chinese Long) - Inclusion
(Chinese Long) - Layout-Overall
(Chinese Long) - Logical Reasoning
(Chinese Long) - Material
(Chinese Long) - Negation
(Chinese Long) - Non Contact
(Chinese Long) - Overall
(Chinese Long) - Pronoun Reference
(Chinese Long) - Quantity
(Chinese Long) - Relationship-Overall
(Chinese Long) - Shape
(Chinese Long) - Similarity
(Chinese Long) - Size
(Chinese Long) - State
(Chinese Long) - Style
(Chinese Long) - Text
(Chinese Long) - World Knowledge
(Chinese) - 3D
(Chinese) - Action-Overall
(Chinese) - Animal
(Chinese) - Attribute-Overall
(Chinese) - Color
(Chinese) - Comparison
(Chinese) - Composition
(Chinese) - Compound-Overall
(Chinese) - Consistency
(Chinese) - Contact
(Chinese) - Expression
(Chinese) - Feature matching
(Chinese) - Full body
(Chinese) - Grammar-Overall
(Chinese) - Hand
(Chinese) - Imagination
(Chinese) - Inclusion
(Chinese) - Layout-Overall
(Chinese) - Logical Reasoning
(Chinese) - Material
(Chinese) - Negation
(Chinese) - Non Contact
(Chinese) - Overall
(Chinese) - Pronoun Reference
(Chinese) - Quantity
(Chinese) - Relationship-Overall
(Chinese) - Shape
(Chinese) - Similarity
(Chinese) - Size
(Chinese) - State
(Chinese) - Style
(Chinese) - Text
(Chinese) - World Knowledge
(English Long) - 2D
(English Long) - 3D
(English Long) - Action-Overall
(English Long) - Animal
(English Long) - Attribute-Overall
(English Long) - Color
(English Long) - Comparison
(English Long) - Composition
(English Long) - Compound-Overall
(English Long) - Consistency
(English Long) - Contact
(English Long) - Expression
(English Long) - Feature matching
(English Long) - Full body
(English Long) - Grammar-Overall
(English Long) - Hand
(English Long) - Imagination
(English Long) - Inclusion
(English Long) - Layout-Overall
(English Long) - Logical Reasoning
(English Long) - Material
(English Long) - Negation
(English Long) - Non Contact
(English Long) - Overall
(English Long) - Pronoun Reference
(English Long) - Quantity
(English Long) - Relationship-Overall
(English Long) - Shape
(English Long) - Similarity
(English Long) - Size
(English Long) - State
(English Long) - Style
(English Long) - Text
(English Long) - World Knowledge
(English) - 2D
(English) - 3D
(English) - Action-Overall
(English) - Animal
(English) - Attribute-Overall
(English) - Color
(English) - Comparison
(English) - Composition
(English) - Compound-Overall
(English) - Consistency
(English) - Contact
(English) - Expression
(English) - Feature matching
(English) - Full body
(English) - Grammar-Overall
(English) - Hand
(English) - Imagination
(English) - Inclusion
(English) - Layout-Overall
(English) - Logical Reasoning
(English) - Material
(English) - Negation
(English) - Non Contact
(English) - Overall
(English) - Pronoun Reference
(English) - Quantity
(English) - Relationship-Overall
(English) - Shape
(English) - Similarity
(English) - Size
(English) - State
(English) - Style
(English) - Text
(English) - World Knowledge
WISE VerifiedSource: WISE data

Human preference

Data credits: LMArena, CC BY 4.0 (licence) · GenExam (OpenGVLab), MIT (licence) · UniGenBench++ (CodeGoat24), MIT (licence)

Also known as: black-forest-labs/flux-2-dev, black-forest-labs/flux.2-dev, flux-2-dev, flux-2-dev-image-to-image, flux-2-dev-lora, flux-2-dev-lora-image-to-image, together_ai/black-forest-labs/flux.2-dev

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works