1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
const axios = require('axios');
const api_key = "YOUR API-KEY";
const url = "https://api.segmind.com/v1/bria-fibo-generate";
const data = {
"prompt": "A moon-kissed rooftop garden in Paris, where strings of golden lights sway like fireflies above lavender and rosemary pots, the Eiffel Tower shimmering in the distance beyond a wrought-iron railing, and a soft breeze carries the quiet rustle of leaves and distant accordion notes into the starlit hush.",
"guidance_scale": 4,
"aspect_ratio": "1:1",
"steps_num": 40,
"seed": 123456,
"ip_signal": false,
"prompt_content_moderation": true,
"visual_input_content_moderation": true,
"visual_output_content_moderation": true
};
(async function() {
try {
const response = await axios.post(url, data, { headers: { 'x-api-key': api_key } });
console.log(response.data);
} catch (error) {
console.error('Error:', error.response.data);
}
})();Attributes
Main instruction for the image creation. Example: 'A serene beach at dawn'.
URL of an input image. Use when refining existing images.
A string containing the structured prompt in JSON format. Use a structured_prompt from a previous generation's response or the /v1/bria-fibo-generate endpoint for precise refinement.
Exclude elements from image. Example: 'No shadows'.
Control prompt adherence. Higher for detailed images.
min : 3,
max : 5
Choose image dimensions. 16:9 for widescreen visuals.
Allowed values:
Diffusion step count. Set higher for smoother results.
min : 20,
max : 50
Use seed for repeatability. Random if omitted.
min : 1,
max : 999999999
Warns about potential IP conflicts. Set true to receive alerts.
Enable to block inappropriate prompts.
Active when blocking inappropriate input images.
Blocks inappropriate image outputs. True if moderation needed.
To keep track of your credit usage, you can inspect the response headers of each API call. The x-remaining-credits property will indicate the number of remaining credits in your account. Ensure you monitor this value to avoid any disruptions in your API usage.
Bria FIBO: Next-Generation Text-to-Image Model
What is Bria FIBO?
Bria FIBO is an advanced AI image generation model designed for photorealistic, high-quality visuals with exceptional prompt accuracy and creative control. Unlike traditional text-to-image models, FIBO uses a two-stage architecture: a Visual Language Model (VLM) Bridge translates natural language or structured JSON prompts into detailed machine-readable instructions, which the FIBO model then uses to generate images. This approach delivers professional-grade results ideal for commercial workflows, marketing campaigns, and scalable content production. FIBO supports Gemini 2.5 Flash for state-of-the-art prompt understanding, with an open-source FIBO-VLM option coming soon for flexible deployments.
Key Features
- Structured JSON Prompts: Achieve precise control over image details using JSON-formatted instructions alongside natural language prompts.
- Dual-Stage Generation: VLM Bridge ensures prompts are interpreted accurately before image creation, reducing ambiguity.
- Image-Inspired Workflows: Refine existing images by providing input URLs for iterative design improvements.
- Aspect Ratio Flexibility: Generate images in nine preset ratios (1:1, 16:9, 9:16, etc.) for diverse use cases.
- Built-In Content Moderation: Multi-layer safeguards block inappropriate prompts, input images, and outputs, ensuring brand-safe results.
- IP Signal Alerts: Optional warnings for potential intellectual property conflicts, useful for risk-aware production environments.
Best Use Cases
FIBO excels in scenarios requiring photorealism, consistency, and legal compliance:
- Marketing and Advertising: Create campaign visuals, product mockups, and hero images with brand-appropriate styling.
- E-commerce: Generate lifestyle photos, product variations, and seasonal assets at scale.
- Media and Publishing: Produce editorial illustrations, book covers, and social media graphics.
- Enterprise Creative Teams: Streamline workflows with repeatable seeds, structured prompts, and refinement capabilities.
Prompt Tips and Output Quality
Writing Effective Prompts:
- Use clear, descriptive language: "A minimalist living room with natural light, beige tones, and indoor plants" outperforms vague instructions.
- Leverage negative prompts to exclude unwanted elements (e.g., "No shadows" or "No text overlays").
- Experiment with structured prompts (JSON) for complex scenes requiring specific object placement, lighting, or composition.
Parameter Impact:
- Guidance Scale (3–5): Higher values (e.g., 5) increase prompt adherence but may reduce creative variation. Start at 4 for balanced results.
- Steps Number (20–50): More steps (e.g., 40+) produce smoother, more refined images but increase generation time.
- Seed: Use consistent seeds for reproducibility across iterations or variations.
- Aspect Ratio: Match output dimensions to your platform (16:9 for web banners, 4:5 for Instagram).
FAQs
Is Bria FIBO open-source?
No, FIBO is a commercial model accessed via API. However, an open-source FIBO-VLM option is planned for flexible deployments.
How does FIBO differ from other text-to-image models?
FIBO's dual-stage architecture (VLM Bridge + image generation) ensures superior prompt accuracy and supports structured JSON inputs for granular control, making it ideal for professional workflows.
What parameters should I tweak for best results?
Focus on guidance_scale (4–5 for detail), steps_num (40+ for quality), and negative_prompt (to exclude unwanted elements). Use seed for repeatable outputs.
Can I refine existing images?
Yes. Provide an image URL via the image_urls parameter to iterate on existing visuals.
What is IP Signal, and should I enable it?
IP Signal warns about potential intellectual property conflicts. Enable it (ip_signal: true) for risk-aware commercial projects.
Does FIBO support content moderation?
Yes. FIBO includes moderation for prompts, input images, and outputs by default, ensuring brand-safe content.
Popular Models
