A Practical Checklist for Evaluating Leonardo.Ai on an Image Task
Research guide · 6 min read
This guide is for someone deciding whether Leonardo.Ai fits a specific image creation workflow. Leonardo’s help center describes image generation, reference inputs, editing, and model-dependent settings. Those vendor descriptions are not evidence of a particular output quality. This article does not report the former personal image test or claim that Leonardo is “the best.”
Set one visual goal
Write a brief with the subject, composition, aspect ratio, style constraints, required elements, prohibited elements, and final use. If your task involves editing, save the original asset and describe exactly what should change and what should remain intact.
Fixed practice task
Example task — not performed by AI Pickz. Use either a rights-cleared product photo that already has the exact label below, or generate a made-up reusable bottle and explicitly request the label. Before editing, save the original and confirm the label reads “FIELD / 750 ml.” Only then open the image in Leonardo’s Inline Editor and change the background to warm ivory. Score the edit only on preserving details that were visibly present in that saved original. Save the prompt, model, references, and every output.
Create a clean square studio image of one matte green reusable water bottle standing upright on a light surface, centered, with a soft shadow cast to the lower right. Put the exact readable label “FIELD / 750 ml” on the bottle. Use a plain pale blue backdrop. No extra objects. Before editing, verify and save an original that visibly contains that exact label. Then change only the backdrop to warm ivory; preserve the bottle silhouette, label, camera angle, placement, and shadow direction. If the generated label is not exact, correct it in the source image or use a rights-cleared source image with that label before evaluating the background edit.
Before starting the edit, inspect the original and confirm that the label visibly reads “FIELD / 750 ml.” If it does not, correct the original or choose a rights-cleared reference image that does; do not count a missing pre-edit label as an edit failure. Score the background change separately from preservation of details that were visibly present in the saved original. This prevents initial text rendering from being confused with edit quality. Leonardo’s current guide describes selecting a generated image, opening Edit/Inline Editor, and giving explicit natural-language instructions; available reference count depends on the selected model.
Record settings and outputs
- Record date, account tier, model, mode, dimensions, references, and token cost displayed before generation.
- Save every output for each prompt and note the exact prompt and settings.
- Test one representative edit such as image guidance, inline editing, or an inpainting change.
- Score adherence to the brief, consistency, artifacts, edit locality, and amount of manual cleanup.
- Check whether Private Mode is available on your plan and is enabled before generating. Leonardo’s Help Center says Private Mode is for paid members; when it is off, generated images are visible in the public feed. Confirm rights for uploaded references and generated assets before client or commercial use.
Track total effort and limits
| Measure | Record |
|---|---|
| Model/version and generation mode | ________________ |
| Prompt and reference assets | ________________ |
| Generation and edit attempts | ________________ |
| Token/credit use and plan cost | ________________ |
| Artifacts and manual corrections | ________________ |
| Privacy and usage terms checked | ________________ |
Sources and verification
Leonardo.Ai official Help Center guides on Omni/Inline Editor, image guidance, image generation, and visibility were checked October 22, 2026. The help center notes that options vary by model and some legacy guides may reflect earlier branding. Check the live app and current terms before relying on a feature.
- Leonardo.Ai Help Center — Generate images
- Leonardo.Ai Help Center — Omni models and inline editor
- Leonardo.Ai Help Center — Image guidance
- Leonardo.Ai Help Center — Private and public images
Limit: AI Pickz did not run this example and has no supplied prompt log, output set, token receipt, or usage-rights review. No image-quality score or subscription recommendation is claimed.