Text to image
Elo · text-to-image arena
FLUX.2 overview
FLUX.2 by Black Forest Labs
Perfect for designers, marketers, and creative teams seeking speed, consistency, and professional-grade results.
FLUX.2 specs
- Model ID
blackforestlabs/flux-2-pro- Developer
- Black Forest Labs
- Input
- Text
- Output
- Image
- Endpoint
POST https://api.aimlapi.com/v1/images/generations
FLUX.2 Pro API Overview
Flux.2 Pro is the premium iteration of the Flux.2 model family, designed for creators and professionals who demand exceptional control, clarity, and consistency in AI‑generated visuals. Unlike general‑purpose models, Flux.2 Pro is built to serve real‑world production needs, maintaining fidelity across high‑resolution outputs, handling complex natural‑language directives, and preserving visual integrity throughout iterative workflows.
Flux.2 Pro Edit expands these capabilities by offering context‑aware transformation tools that go beyond simple alterations, allowing precise refinement of existing visuals with language‑based instructions that feel intuitive yet meticulously accurate.
Performance Benchmarks
- Matches or exceeds quality levels of many closed-source state-of-the-art models while operating faster and at a lower cost.
- Demonstrates stable and predictable output quality, reducing manual tuning overhead.
What Sets Flux.2 Pro Apart?
Whether generating product images, marketing assets, or campaign visuals, the model interprets prompts with profundo semantic understanding and returns outputs at up to 4 megapixels of detail, robust enough for high‑end campaigns and large‑format use. The underlying design supports multi‑image reference input, meaning users can anchor style, lighting, layout, and character consistency across every image produced. This empowers teams to maintain a unified visual language throughout a series of assets without repetitive manual refinement.
Flux.2 Pro’s proficiency extends to natural language editing as well. Instead of traditional editing interfaces with layers and masks, users describe desired changes from swapping backgrounds to adjusting garment color and the model executes transformations while preserving original context and visual logic.
Flux.2 Pro Edit: Intelligent, Context‑Driven Image Refinement
Flux.2 Pro Edit takes Flux.2 Pro’s generative strength and applies it specifically to editing tasks that require precision and nuance. Rather than applying generic effects, the model interprets the structure and meaning of a visual scene, allowing creators to refine images in ways that feel both creative and deliberate.
This means you can evolve an existing image step by step, improving composition, refining lighting, enhancing realism, or fine‑tuning elements like text and layout, all driven by clear, natural instructions. Flux.2 Pro Edit reduces the friction of traditional editing while amplifying expressive control.
From subtle adjustments to more ambitious revisions, this workflow supports both iterative experimentation and professional polish.
Why Professionals Choose Flux.2 Pro
Flux.2 Pro is designed for users who need quality without compromise. It supports brand consistency across batches of assets, preserves character and design continuity, and renders textual elements, such as product labels, banners, and headlines with remarkable clarity.
The result is a tool that feels both powerful and responsive, capable of producing studio‑grade visuals without the steep learning curve of traditional creative software. This makes Flux.2 Pro particularly valuable for advertising teams, content studios, product designers, UX/UI creators, and anyone who relies on compelling imagery as a core part of their workflow.
Use cases
- Professional content creation pipelines requiring high-quality generative imagery.
- Automated image generation for marketing, e-commerce product visuals, and brand-consistent materials.
- Workflows where batch consistency and reproducibility are critical.
- Scenarios where minimal technical tuning is preferred to accelerate deployment speed.
FLUX.2 benchmarks by 
Artificial Analysis
Human-preference results for FLUX.2 in text to image, next to models available on AI/ML API. Data retrieved Nov 3, 2026.
AI/ML API price
USD per Megapixel
Models on AI/ML API
Arena results for models you can call with the same API key.
Text to image
Blind pairwise human votes on images generated from text prompts. FLUX.2 ranks #44 of 98.
FLUX.2 API pricing
| Usage | Price | Unit |
|---|---|---|
| Image generation | $0.03 | per Megapixel |
The final price depends on request parameters such as resolution or duration.
Prices from the AI/ML API price list, checked on Nov 4, 2026.
FLUX.2 vs other image models: prices
AI/ML API rates for 5 models, per Megapixel.
| Model | Per Megapixel |
|---|---|
 FLUX.2 | $0.03 |
| FLUX.2 Klein 4B | $0.014 |
| FLUX.2 Klein 9B | $0.015 |
| FLUX.2 Klein 9B Preview | $0.015 |
| FLUX.2 Flex | $0.05 |
How to use the FLUX.2 API
View docs
Build with your agent
Give your agent the model details, code and documentation.
Paste it into your coding agent to build the integration.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v1/images/generations' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "blackforestlabs/flux-2-pro",
"prompt": "A ginger cat asleep on a stack of exactly three hardcover books in a rainy bookshop window at dusk. Gold lettering on the glass reads OPEN LATE. Warm lamp light inside, raindrops and blurred neon reflections outside, 35mm film photo."
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "blackforestlabs/flux-2-pro",
"prompt": "A ginger cat asleep on a stack of exactly three hardcover books in a rainy bookshop window at dusk. Gold lettering on the glass reads OPEN LATE. Warm lamp light inside, raindrops and blurred neon reflections outside, 35mm film photo."
}
response = requests.post("https://api.aimlapi.com/v1/images/generations", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v1/images/generations", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "blackforestlabs/flux-2-pro",
"prompt": "A ginger cat asleep on a stack of exactly three hardcover books in a rainy bookshop window at dusk. Gold lettering on the glass reads OPEN LATE. Warm lamp light inside, raindrops and blurred neon reflections outside, 35mm film photo."
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());Models similar to FLUX.2
Related articles
.webp)
FLUX.2 FAQ
What inputs and outputs does FLUX.2 [pro] support?
FLUX.2 [pro] takes image, text as input and returns image.
How much does FLUX.2 [pro] cost?
FLUX.2 costs $0.03 per Megapixel on AI/ML API, billed as you go with no subscription.
How is FLUX.2 [pro] billed?
FLUX.2 [pro] is billed per generation, a fixed charge per output rather than by prompt length.
Can FLUX.2 [pro] take an existing image as input?
Yes, FLUX.2 [pro] accepts image input as well as a text prompt.
Who created FLUX.2 [pro]?
FLUX.2 [pro] was built by Black Forest Labs.
How do I call FLUX.2 [pro] on AI/ML API?
Send a request to https://api.aimlapi.com/v1/images/generations with flux-2-pro as the model id.
Is FLUX.2 [pro] available through a single API key?
Yes. FLUX.2 [pro] is served through AI/ML API, so the same key and endpoint format used for other models applies.
What is the model identifier for FLUX.2 [pro] when calling the API?
The model id is flux-2-pro.