Black Forest Labs: FLUX.2 [flex]

FLUX.2 Flex gives image creators control over how closely generation follows a prompt. Its emphasis on typography and text rendering makes it useful for posters, packaging concepts, and layouts that combine words with imagery.

Compare models

flux-2-flex

AvailableBlack Forest LabsImageSync / Async

Price: From $0.05
Released: Nov 25, 2025
Modalities: Image

About FLUX.2 [flex]

FLUX.2 [flex] is the FLUX.2 variant from Black Forest Labs that exposes generation controls and targets text and typography. It creates images from text and edits them using up to eight reference images. Unlike the other FLUX.2 API models, it lets you set the guidance scale and step count, trading speed against detail and prompt adherence.

Where it works well

Black Forest Labs built it for text and typography rendering while keeping small details intact, which fits signs, labels, and layouts.
Guidance scale (1.5 to 10) and step count (1 to 50) are adjustable, so you can tune adherence against realism per request.
It accepts up to eight reference images for multi-image editing and composition.

When to choose another model

Higher guidance improves prompt adherence but reduces realism, so tuning is required; the fixed-setting FLUX.2 [pro] is simpler when you do not want to tune.
Using the full step count gives the most detail but takes longer than the speed-focused Klein models.
Black Forest Labs lists it as a FLUX.2 model, not the top-end quality tier; FLUX.2 [max] targets maximum fidelity.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Image editingText to imageTokenLab endpoint

POST/v1/images/generations

curl -X POST "https://api.tokenlab.sh/v1/images/generations" \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "operation": "text-to-image",
  "model": "flux-2-flex",
  "prompt": "A minimalist product photo of matte black headphones on a soft blue background."
}'

Parameters

Image editing

Parameter	Values and limits	Default	Notes
`async`	Not provided	Not provided	Not provided
`guidance`	≥ 1.5, ≤ 10	Not provided	Guidance from 1.5 through 10. The model default is not declared.
`input_image`	Inputs: ≥ 1, ≤ 1	Not provided	One source-image URL. Additional source-image fields are not declared as configurable public parameters.
`operation`	Not provided	Not provided	Not provided
`prompt`	Not provided	Not provided	Not provided
`resolution`	Not provided	Not provided	Use explicit pixel dimensions such as 1024x1024. The verified pixel contract has a minimum edge of 64, a 16-pixel step and at most 4194304 total pixels. Tier labels such as 1k or 2k are not a declared finite resolution list.

Text to image

Parameter	Values and limits	Default	Notes
`async`	Not provided	Not provided	Not provided
`guidance`	≥ 1.5, ≤ 10	Not provided	Guidance from 1.5 through 10. The model default is not declared.
`input_image`	Not provided	Not provided	Text-to-image does not accept source images. Do not send this field for that operation. Use one of this model's declared image operations for source-image requests.
`operation`	Not provided	Not provided	Not provided
`prompt`	Not provided	Not provided	Not provided
`resolution`	Not provided	Not provided	Use explicit pixel dimensions such as 1024x1024. The verified pixel contract has a minimum edge of 64, a 16-pixel step and at most 4194304 total pixels. Tier labels such as 1k or 2k are not a declared finite resolution list.

image: Text-to-image does not require a source image. Source-image controls are used by the declared image operations.

Task-capable image contract; `async: true` is accepted when the public status mode is `inline_or_task`.

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

Pricing

per image

Verified price: —
Official price: How pricing works

	Official priceper image	Verified priceper image	Discount
Price	How pricing works	—	—

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Nov 1, 2026, 6:32 AM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open FLUX.2 [flex] in Console with a prompt ready to edit or send.

Help me create an image with flux-2-flex using text-to-image at /v1/images/generations. Show the result, cost, and any limits I should know.

Use cases

Best for

Text to Image

Posters and packaging mockups
Set type-heavy briefs, such as a poster headline or a product label, and raise guidance until the lettering follows the prompt.
Multi-reference composition
Combine a product shot, a style reference, and a background image into one scene using several input images.
Iterating on a layout
Re-run the same seed with different step counts or guidance values to find the setting that balances detail and speed.

Prompt examples

A concert poster with the headline 'NIGHT MARKET' in bold condensed type, neon pink on deep blue, small credits along the bottom.

Cereal box front, flat vector style, title 'Honey Oat Crunch' in rounded letters, a smiling bee mascot.

Edit the attached storefront photo: change the sign to read 'Blue Door Books' in white serif capitals.

FAQ

What makes FLUX.2 [flex] different from FLUX.2 [pro]?

[flex] exposes guidance scale and step count, and is aimed at text and typography. [pro] runs with fixed settings for a simpler workflow. Choose [flex] when lettering accuracy or per-request tuning matters more than convenience.

How many reference images does FLUX.2 [flex] accept?

Up to eight. Black Forest Labs documents input_image through input_image_8 for editing and composition, alongside the text prompt, so one request can merge several sources into a single result.

What do guidance and steps do on FLUX.2 [flex]?

Guidance (1.5 to 10, default 5) sets how closely the model follows the prompt; higher values stick to it more but look less realistic. Steps (1 to 50, default 50) set the detail level; fewer steps are faster.

Is FLUX.2 [flex] good for text in images?

Yes, that is its stated focus. Black Forest Labs describes it as specialized for text and typography while preserving small details. Spell out the exact words in quotes and keep them short for the most reliable results.

How much does FLUX.2 [flex] cost?

Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

Which endpoint should FLUX.2 [flex] use?

Use https://api.tokenlab.sh/v1/images/generations for FLUX.2 [flex]. The request example below shows the matching code shape.

Which operations does FLUX.2 [flex] support?

FLUX.2 [flex] supports Image editing, Text to image. Select an operation above to see its endpoint and request example.