FLUX.2 Flex
FLUX.2 [flex] is the tunable tier of Black Forest Labs' FLUX.2 family, released on 25 November 2025. Where [pro] hides the diffusion controls, [flex] exposes the number of inference steps and the guidance scale, and BFL pitches it for typography and for trading speed against quality on purpose. It costs five cents a megapixel.
Anvisha Pai, Co-founder & CEO, ModaTested Black Forest Labs · released
›What we asked for
A self-portrait of FLUX.2 Flex, Black Forest Labs' tunable image model, as a detailed pencil and watercolour illustration: an old brass control panel with dials labelled 'STEPS' and 'GUIDANCE', a hand adjusting one, loose washes of indigo and ochre bleeding past the linework, sketchbook margins.
fal-ai/flux-2-flex
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
At a glance
- Price
- $0.05 per megapixel, so a 1K image is five cents.
- Steps
- How many passes the model takes, default 28. The ladder below shows 10, 28 and 50 on the same prompts, with time and cost.
- Guidance
- How hard it sticks to the prompt, default 3.5.
- Typography
- BFL's stated strength for this tier; the poster and infographic samples are the test.
- Any size
- Exact width and height, priced by area.
Other FLUX models: FLUX.2 Pro, FLUX.2 Max. Compare all 3.
Anvisha's take
Flex is Pro with the hood open, and the ladder is the reason to read this page. At 10 steps the poster came back with the type piled on itself; at 28 and 50 it was clean and correct. The pizza barely changed. Steps cost nothing extra on fal, only seconds, so the default of 28 is the floor, not a suggestion. The infographic was tidy and complete, and I would use Flex over Pro whenever the picture has words in it.
FLUX.2 Flex real output: the same 12 prompts every model gets
Every model on this site gets the same 12 prompts, so you can compare like with like. These are the pictures FLUX.2 Flex returned, one try each, nothing re-rolled or retouched. Generated on . The 12 pictures took 9 to 13 seconds each and cost $1.05 in total.
Spelling, kerning and hierarchy of three lines of set text.
›What we asked for
A concert poster that reads 'MODA SUMMER SESSIONS' in large bold type, with the subtitle 'Friday 12 June, 8pm, Roundhouse' and a smaller line 'Tickets at the door'. Risograph style, two colours, tangerine and navy, grainy ink texture.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Catalogue lighting, a clean seamless background and a believable contact shadow.
›What we asked for
Studio product photograph of a matte black ceramic pour-over coffee dripper on a seamless white background, soft top light, subtle contact shadow, centred, catalogue style.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
A natural face, correct finger count and hands holding an object.
›What we asked for
Close-up portrait of a woman in her 40s with short grey hair, laughing, holding a red ceramic mug with both hands, natural window light, 85mm lens, shallow depth of field.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
A flat single-colour mark with a legible wordmark and no gradients or photo texture.
›What we asked for
A flat vector logo mark for a bicycle repair shop called 'Spoke & Chain': a single-colour icon combining a chain link and a wheel spoke, with the wordmark below in a geometric sans-serif, black on white.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Many distinct objects and people in one frame without merging or duplicating them.
›What we asked for
A crowded Saturday farmers market seen from slightly above: at least twelve stalls with striped awnings, shoppers carrying tote bags, a dog on a leash, crates of oranges, a musician with a guitar, bunting overhead, late morning light.
fal-ai/flux-2-flex
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
A named historical style applied faithfully rather than a generic painterly look.
›What we asked for
A lighthouse on a rocky cliff during a storm, in the style of a Japanese ukiyo-e woodblock print, with visible wood grain texture, flat colour areas and a bokashi gradient sky.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Six labelled bars at the stated heights, in order, with readable axes.
›What we asked for
A clean bar chart titled 'Monthly active users' showing six months, January through June, with the values 120, 150, 180, 170, 220 and 260, labelled axes, white background, one accent colour, no 3D effects.
fal-ai/flux-2-flex
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
A 21:9 composition that leaves the left third clear for headline text.
›What we asked for
Website hero banner: a wide panorama of a sunlit modern co-working space with plants, the left third of the frame kept as clear wall space for text, warm neutral palette.
fal-ai/flux-2-flex
image_size: {"width":2016,"height":864}, output_format: png, num_images: 1
returned 2016 by 864 pixels
Four numbered steps in the given order with their exact wording, plus an icon each.
›What we asked for
An infographic card titled 'How to brew pour-over' with four numbered steps in this order: 1. Rinse the filter, 2. Bloom for 30 seconds, 3. Pour in slow circles, 4. Wait 3 minutes. A simple icon beside each step, muted palette, clean sans-serif type.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1360}, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Appetising, menu-grade food with correct textures and no melted anatomy.
›What we asked for
Overhead photograph of a margherita pizza on a wooden board, fresh basil leaves, one slice pulled out with the cheese stretching, restaurant menu photography, natural light.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Exact counts and left-right placement of five named objects.
›What we asked for
A glass of water on a wooden table with exactly three ice cubes in it, a yellow pencil resting across the rim of the glass, a blue book to the left of the glass and a red apple to the right, photographed from table height.
fal-ai/flux-2-flex
image_size: {"width":1360,"height":1024}, output_format: png, num_images: 1
returned 1360 by 1024 pixels
Change one object's colour and add a part while leaving background, light and framing untouched.
›What we asked for
Change the coffee dripper to a deep forest green and add a small brass handle on its right side. Keep the background, lighting and composition exactly as they are. (applied to the product on white result above)
fal-ai/flux-2-flex/edit
image_size: {"width":1024,"height":1024}, output_format: png, num_images: 1
returned 1024 by 1024 pixels
How the cost under each picture was worked out: fal's FLUX.2 [flex] rate of $0.05 per megapixel, rounded up to the nearest 1024x1024 megapixel as fal states: 2 for the 1024x1360 output; fal does not return the charge.
FLUX.2 Flex settings that matter
Inference steps
The number of steps changed the poster far more than the pizza. At 10 the headline collapsed into overlapping type; at 28, the default, everything was correct; at 50 it was cleaner still with smaller type. The pizza was fine at all three. Since fal charges the same at any step count, the only cost is seconds: 7, 8 and 11 on the poster. Never go below the default for text; go above it when the layout is complex.
Text in an image
›What we asked for
A concert poster that reads 'MODA SUMMER SESSIONS' in large bold type, with the subtitle 'Friday 12 June, 8pm, Roundhouse' and a smaller line 'Tickets at the door'. Risograph style, two colours, tangerine and navy, grainy ink texture.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1360}, num_inference_steps: 10, output_format: png, num_images: 1
returned 1024 by 1360 pixels
Food photography
›What we asked for
Overhead photograph of a margherita pizza on a wooden board, fresh basil leaves, one slice pulled out with the cheese stretching, restaurant menu photography, natural light.
fal-ai/flux-2-flex
image_size: {"width":1024,"height":1024}, num_inference_steps: 10, output_format: png, num_images: 1
returned 1024 by 1024 pixels
FLUX.2 Flex pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| BFL API | 1K image Priced by the megapixel, so this is the starting rate for a 1K image. | $0.05 | BFL pricing, read |
| fal | 1024 by 1024 image $0.05 per megapixel of output, and per megapixel of input on edits. | $0.05 | fal model page, read |
Is FLUX.2 Flex free?
Black Forest Labs sells FLUX.2 by the megapixel through its own API and partners and does not run a free consumer app. Free access exists only through third parties: hosts and creative apps that bundle FLUX with sign-up credits, and the open-weight FLUX.2 [dev] and [klein] models you can run yourself.
- •BFL API: pay per megapixel, no free tier. Batches multiply the base price by the image count.
- •Open weights: FLUX.2 [dev] and [klein] can be downloaded and run locally under their licences; [klein] is Apache 2.0.
- •Third-party apps: sites like Krea, Freepik and Mistral's Le Chat bundle FLUX models with their own free allowances.
BFL pricing, read
Where to use FLUX.2 Flex
- BFL API · Black Forest Labs' own endpoint, flux-2-flex.
- fal · Per-megapixel pricing, generate and edit endpoints, custom sizes.
- Replicate · Hosted under black-forest-labs.
- Moda · Available inside the Moda design agent.
Generate with FLUX.2 Flex in Moda
Pick FLUX.2 Flex in the Moda design agent, then edit the result as vectors, text and layers rather than a flat image.
Open Moda →What we noticed running FLUX.2 Flex
- •Four headings, exact wording, an icon each, and no invented copy, in a clean card layout. Better designed than Pro's on the same prompt. See the infographic with steps picture.
- •Text is correct at the default steps; see the ladder for what happens below it. See the text in an image picture.
- •Steps cost time, not money. fal prices Flex by the megapixel, so 10, 28 and 50 steps all billed the same, $0.10 for the 1024 by 1360 poster, which rounds up to two megapixels; they took 7, 8 and 11 seconds.
- •Fast: 9 to 13 seconds per picture on the fixed set.
- •Draws its own controls: the self-portrait rendered dials labelled STEPS and GUIDANCE with plausible scales, which is either charming or a warning about how literally it reads prompts.
- •Custom sizes are honoured to the pixel.
FLUX.2 Flex limits and API parameters
| Limit | Value |
|---|---|
| Output size | Any width and height, priced by the megapixel; our 21:9 run at 2016 by 864 came back at that exact size. fal API reference, read |
| Transparent background | Not supported on the hosted endpoints. fal API reference, read |
| Reference images | Multiple, through the edit endpoint; BFL cites up to 10 for [flex]. FLUX.2 announcement, read |
›Every setting, for developers
fal (fal-ai/flux-2-flex)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Any text Required. | |
| image_size | enum or object | square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9, or {width, height} | landscape_4_3 |
| seed | integer | Any integer | |
| safety_tolerance | enum | 1 to 5, 1 strictest | 2 |
| enable_safety_checker | boolean | true, false Turning it off needs account authorisation. | true |
| output_format | enum | jpeg, png | jpeg |
| guidance_scale | number | Prompt adherence | 3.5 |
| num_inference_steps | integer | Diffusion passes | 28 |
| image_urls | list of URLs | Reference images Edit endpoint only, where it is required. |
What FLUX.2 Flex will not draw
Black Forest Labs licenses the FLUX.2 API models for commercial use and ships them with a safety checker that hosts leave on by default; fal exposes a safety_tolerance dial from 1 to 5, 1 strictest, and turning the checker off entirely needs account authorisation. The usual prohibitions apply: sexual content involving minors, non-consensual intimate imagery, and deceptive depictions of real people. There is no adult mode on the hosted models.
fal API reference, read
Alternatives to FLUX.2 Flex
Frequently asked questions
What is FLUX.2 Flex?
FLUX.2 [flex] is the tier of Black Forest Labs' FLUX.2 family that exposes inference steps and guidance scale, released on 25 November 2025 and priced at $0.05 per megapixel.
How many steps should I use?
The ladder above runs the same two prompts at 10, 28 and 50 steps. 28 is the default; look at whether 50 changes anything you can see before paying the extra time.
Flex or Pro?
Pro for photos and volume. Flex when you want to push typography or tune the trade-off between speed and detail yourself.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
