Image Generation Documentation

Generate images across Replicate, Vertex AI, OpenAI, and Fireworks with prompt suggestions, provider-level controls, and rich history.

Overview

Provider routing

Replicate, Vertex AI, OpenAI, and Fireworks models live in the same panel with per-provider defaults.

Prompt controls

Suggestion chips, credit warnings, provider/model selectors, and persistent settings keep runs consistent.

History & telemetry

Review render times, failures, and download links from the `/image/history` gallery.

Replicate

Run Stability, Flux, and community checkpoints with full model ID control and timing logs.

Vertex AI

Use Imagen models for photoreal shots with Google Cloud latency and regional routing.

OpenAI

Generate DALL·E 3 images with built-in safety filters and credit tracking.

Fireworks

Tap flux-1-schnell for ultra-fast ideation or flux-1-dev for higher fidelity.

Select a provider

Pick the model that matches your style, latency, or budget needs.

Craft the prompt

Use suggestion chips, set aspect ratio, and add negative prompts if needed.

Review results

Compare outputs, download originals, or rerun with tweaks.

Manage history

Return to `/image/history` to re-download, filter by provider, or debug failures.

Create your next visual

Route prompts to the best provider, monitor credits, and keep every render in history.