Flux Schnell Text to Image API

blackforestlabs/flux/schnell
png / jpeg

Flux Schnell Text to Image transforms natural language prompts into high-quality visual imagery with 1–4 step ultra-fast inference, native typography rendering, and versatile canvas proportions. It maintains faithful prompt adherence and crisp spatial composition while delivering sub-second response times for high-volume creative iteration.

Input
OutputWaiting

Your images will appear here

Submission charge: 1 MP × 1 × 0.48 = 0.48 credits · $0.0024

0.48 credits per billable MP × image count; 1 credit = $0.005.

Examples

Frost in the Cabbage Patch

A low-angle close-up nature photograph of one purple cabbage still rooted in a quiet outdoor vegetable bed on a frosty morning. Its broad curled outer leaves spread around a compact central head. Delicate white frost crystals trace the ruffled leaf edges and branching purple-green veins, with tiny melt droplets where soft early sunlight reaches the leaf. Show convincing waxy leaf texture and a small amount of dark garden soil below; other plants dissolve into a gentle natural background blur. Rich but realistic violet and muted green, precise near detail, shallow photographic depth of field, calm winter morning light. The cabbage is growing in the earth, not cut or arranged as a product. No people, packaging, text, brands, advertising or watermark.

Kites Above the Hillside Village

A full-bleed horizontal illustration in the style of a modern European linocut. A cheerful kite festival in a small hillside village: three paper kites of visibly different shapes and sizes fly above the houses, their thin strings reaching toward tiny villagers along a winding grassy path. One kite is long and narrow, one has broad triangular wings, and one is a small diamond. Bold indigo blue, ochre yellow and brick red inks on warm cream, hand-carved gouge marks, slightly uneven ink texture, simple readable silhouettes, airy sky and a lively rural composition. The drawing fills the entire image edge to edge. Pure landscape artwork only: every corner contains only sky or landscape, absolutely no calligraphy, writing, signature, stamp, lettering, labels, border, advertisement, logo or watermark.

Flux Schnell Text to Image

Flux Schnell Text to Image is a 12-billion-parameter rectified flow transformer developed by Black Forest Labs, engineered specifically for ultra-fast, production-ready image synthesis. With advanced latent flow-matching distillation, it completes high-fidelity image generation in just 1 to 4 steps, delivering sub-second visual generation with native typography rendering, versatile canvas dimensions, and an Apache 2.0 commercial-friendly license. At 0.48 credits ($0.0024) per megapixel, it provides cost-effective batch generation backed by automatic refunds on uncompleted tasks.

Why Choose Flux Schnell

  • 1–4 Step Ultra-Fast InferenceUses specialized latent guidance distillation to synthesize high-quality images in just 1 to 4 steps, dramatically accelerating creative cycles.

  • 12B Rectified Flow ArchitectureCombines a 12-billion-parameter flow-matching foundation with advanced text-image alignment for coherent composition and rich visual detail.

  • Native Typography & Text GenerationAccurately renders legible in-scene words, signage, and branding by simply placing the target text inside quotation marks in your prompt.

  • Apache 2.0 Commercial-Friendly LicenseReleased under the open Apache 2.0 license, allowing full commercial utilization for marketing, packaging, and enterprise products.

  • Ultra-Low Cost at 0.48 Credits/MPPriced transparently at 0.48 credits ($0.0024) per megapixel with automatic full refunds if an asynchronous generation task encounters an error.

Parameters

ParameterRequirementDescription
promptRequired

Image generation prompt; at least one character after trimming.

sizeOptional

Text-to-image canvas; custom WIDTHxHEIGHT uses positive integer dimensions.

Default1:11:1 HD4:33:416:99:16WIDTHxHEIGHT
nOptional

Positive integer specifying the number of output images.

Default1
output_formatOptional

Output image format.

Defaultpngjpeg

How to Use

  1. Compose the Visual PromptDescribe your subject, scene environment, lighting, and style. Enclose any desired in-image text in quotation marks.

  2. Select Canvas DimensionsChoose a preset aspect ratio such as 1:1, 16:9, or 9:16, or define custom WIDTHxHEIGHT dimensions in positive integer pixels.

  3. Configure Quantity & FormatSet n to your desired image count and select either lossless PNG or compressed JPEG encoding.

  4. Submit Task & Monitor StatusSend a POST request with your API key to /api/generate/submit, obtain the task_id, and poll status until it completes.

  5. Retrieve & Download ImagesRead the generated image URLs directly from data.files[].file_url once the status field transitions to finished.

Pricing

Submission charge: 0.48 credits ($0.0024) per billable MP. Round the selected canvas up to a whole MP per image, with a minimum of 1 MP, then multiply by n. Failed tasks are refunded.

UsageRateDetails
Rate0.48 credits/MP · $0.0024/MP1 MP = 1,000,000 pixels; calculated per image.
1:1 · 4:3 · 3:4 · 16:9 · 9:160.48 credits/image · $0.0024/imageEach preset rounds up to 1 billable MP.
1:1 HD · 1024x15360.96 credits/image · $0.0048/imageEach canvas rounds up to 2 billable MP.
Custom sizemax(1, ceil(W × H / 1,000,000)) × n × 0.48 creditsUse the submitted size and image count.

Use Cases

  • High-Speed Ideation & Rapid PrototypingExplore creative directions, moodboards, and design concepts in real time with near-instantaneous 4-step generation.

  • Batch Social Media & Marketing VisualsGenerate large volumes of visual content and marketing graphics cost-effectively under the Apache 2.0 commercial license.

  • Low-Latency Application IntegrationEmbed rapid visual synthesis into interactive web apps, conversational agents, and real-time generation pipelines.

  • E-commerce Product Mockups & VariationsQuickly generate diverse studio backgrounds, lifestyle settings, and contextual showcases for commercial products.

Tips

  • Enclose any text you want rendered in the image inside quotation marks, specifying its physical placement such as a neon sign or coffee cup label.
  • Structure prompts clearly by describing the main subject first, followed by scene setting, lighting ambiance, and artistic medium.
  • Select your preferred aspect ratio before finalizing the prompt to help the model plan natural spatial framing and balance.
  • Use positive descriptive keywords for textures, materials, and colors rather than relying heavily on negative phrasing.
  • Take advantage of the sub-second generation speed and low cost by generating multiple variants with n to pick the best composition.

Notes

  • Prompt requirement: prompt is required and must contain at least one valid character after trimming whitespace.
  • Megapixel rounding: canvas dimensions (width × height) round up to whole megapixels, with a minimum charge of 1 MP per image.
  • Supported formats: output_format accepts png and jpeg; requests without this parameter default to lossless png.
  • Automatic refund protection: if a generation task encounters a system error or fails to finish, all deducted credits are fully refunded.

Related Models

Flux Schnell Text to Image API frequently asked questions

What is the Flux Schnell Text to Image API?

Flux Schnell Text to Image is a Black Forest Labs model for ultra-fast image generation from natural language descriptions. Flux Schnell Text to Image generates high-fidelity images in 1 to 4 steps with sub-second response times, native typography rendering, and flexible preset or custom canvas dimensions. Built on a 12-billion-parameter rectified flow transformer architecture with latent guidance distillation, it maintains faithful prompt adherence and coherent composition while dramatically reducing inference latency. You can call it programmatically or try it from the playground above.

How many inference steps does Flux Schnell Text to Image use?

Flux Schnell Text to Image operates in just 1 to 4 inference steps, with 4 steps being the optimal standard. By utilizing advanced flow-matching distillation, the model achieves high visual quality without requiring the dozens of iterative denoising steps common in traditional diffusion models.

What aspect ratios and canvas sizes does Flux Schnell Text to Image support?

Flux Schnell Text to Image supports six preset aspect ratios: 1:1 (512×512), 1:1 HD (1024×1024), 4:3 (1024×768), 3:4 (768×1024), 16:9 (1024×576), and 9:16 (576×1024). You can also specify custom WIDTHxHEIGHT dimensions using positive integer pixels to fit any custom design or digital banner requirements.

How does Flux Schnell Text to Image render text within images?

To render specific words or phrases in the generated image, place the exact English text inside quotation marks within your prompt. You can guide typography style by specifying physical materials and scene context, such as a glowing neon sign, an engraved metal plaque, or printed packaging.

What is the pricing rate for Flux Schnell Text to Image?

Generation is billed at 0.48 credits ($0.0024) per billable megapixel (MP), rounded up to the nearest whole megapixel per image. Standard presets including 1:1, 4:3, 3:4, 16:9, and 9:16 cost 0.48 credits ($0.0024) per image, while 1:1 HD costs 0.96 credits ($0.0048). If an asynchronous task fails, all deducted credits are automatically refunded.

Can images generated by Flux Schnell Text to Image be used commercially?

Yes. Flux Schnell is released under the open Apache 2.0 license, which permits both personal and commercial use. You can freely integrate generated visual assets into commercial advertising, digital products, marketing media, and client deliverables.

How does Flux Schnell Text to Image compare to Flux Dev?

Flux Schnell is engineered for ultra-fast generation (1 to 4 steps), lower latency, ultra-low cost (0.48 credits/MP), and commercial use under Apache 2.0. In contrast, Flux Dev uses a 50-step guidance process tailored for intricate fine-grain micro-textures and non-commercial research, making Schnell the ideal choice for high-volume production pipelines.

Does Flux Schnell Text to Image support batch image generation?

Yes. You can generate multiple images from a single prompt by setting the n parameter to an integer greater than zero. The task cost scales linearly by multiplying the single-image MP cost by n, and all completed image URLs are returned together in the data.files array.