FIBO

Released
2025-11-03
Last refreshed
2026-05-19
Status
Researched 174d ago
Open weightsCommercial use: not permittedMultimodal

FIBO is a released general LLM work model with open-weight; evaluate it while provider pricing coverage matures.

BRIA releases · 2 in the last 12 monthsChangelog →
Specifications
Family
FIBO
Released
2025-11-03
Parameters
8B
Architecture
Diffusion Transformer
Specialization
image-generation
Openness
Open weights
License
CC-BY-NC-4.0Not permitted
Weights
Available
Code
Unknown
Created by
BRIA

AI-powered visuals, responsibly scaled

Tel Aviv, Israel
Founded 2020
Website
Pricing

No tracked provider pricing is available yet.

Links
Website HuggingFace

About

FIBO by Bria AI. The world's first open-source JSON-native text-to-image model (8B DiT with flow-matching). Trained exclusively on licensed, multi-domain data. Uses structured JSON captions (~1,000 words, 100+ visual attributes) for deterministic, brand-safe image generation. Supports iterative refinement of individual image attributes. Available on GitHub (Bria-AI/FIBO) and HuggingFace (briaai/FIBO). Integrates with Adobe, Figma, and AWS.

Providers

No active provider routes are tracked for this model yet.

Capabilities

Multimodal

Benchmark peer barsfor Coding

No task-mapped benchmark peers are available for this model yet.

Migration checks

No linked migration route is available for this model yet.