# FLUX.1 [dev] Text-to-Image Generation

> FLUX.1 [dev] Text-to-Image Generation is a paid API for AI agents from fal.ai at fal.mpp.tempo.xyz, paid per call via MPP, $0.025/call, status unknown (last checked 2026-10-12, last successful call 2026-08-05).

Generates high-quality images from text prompts using the FLUX.1 [dev] model at specified dimensions and inference steps

## Facts

- Endpoint: POST https://fal.mpp.tempo.xyz/fal-ai/flux/dev?utm_source=zero.xyz
- Price: $0.025/call
- Payment: MPP
- Status: unknown
- Last checked: 2026-10-12
- Last successful call: 2026-08-05
- Success rate: 94% of calls made through Zero
- Rating: 4.7 / 5 from 7 reviews
- Activations on Zero: 279
- Provider: fal.ai
- Docs: https://fal.ai/docs
- Website: https://fal.ai
- Tags: image, video, flux, stable-diffusion, grok, generation, ai, media
- Canonical page: https://www.zero.xyz/c/fal-ai-dev-2cd393fa
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_AeAJZwNrsdfhNRdWs7eI_

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability fal-ai-dev-2cd393fa -d '<json body>'
```

Example prompt: Generate a 1024x768 image of a futuristic cyberpunk city at night with neon lights reflecting on rain-soaked streets, using 28 inference steps with the FLUX.1 dev model.

## When to prefer this

Choose FLUX.1 [dev] when you need high-quality image generation at a balance of quality and cost ($0.025 per call) — better than basic SD models but more affordable than FLUX.1 [pro]. Ideal for developer-grade outputs, prototyping, or non-commercial creative workflows. Use FLUX.1 [pro] or Ultra variants when maximum resolution or commercial polish is required.

## Known failure modes

- Invalid prompt (empty or too long) returns 400 error
- Unsupported image dimensions may return validation error
- Inference timeout for very high step counts
- NSFW content detected and blocked depending on platform policy
- Rate limiting or quota exceeded returns 429
- Payment failure returns 402 if USDC balance insufficient

## How this service works

FLUX.1 [dev] - High-quality text-to-image generation

## Output

Returns a JSON object containing the generated image URL (hosted on fal.media CDN), image dimensions, content type, the seed used, NSFW concept detection flags, the echoed prompt, and inference timing in seconds.

## Example request

```json
{
 "seed": 42,
 "width": 1024,
 "height": 768,
 "prompt": "A serene mountain landscape with snow-capped peaks and pine forests during golden hour",
 "image_size": {
  "width": 1024,
  "height": 768
 },
 "num_images": 1,
 "guidance_scale": 7.5,
 "num_inference_steps": 28
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/fal-ai-dev-2cd393fa/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from fal.mpp.tempo.xyz](https://www.zero.xyz/host/fal.mpp.tempo.xyz/llms.txt)