Leonardo AI Pricing 2026: What Do the Plans and API Actually Cost?
Leonardo AI has two distinct cost questions: what a person pays for the creative app, and what a developer pays to call its API. The app sells monthly plans with token allowances and selected relaxed-generation modes. The API is moving to pay-as-you-go billing, where model and request settings determine cost. A consumer subscription is not an API wallet, and a token is not a fixed number of images.
As checked September 26, 2026, Leonardo lists Free at $0, Essential at $12/month, Premium at $30/month and Ultimate at $60/month, before tax. Its pricing page advertises up to 20% off with yearly billing; those prices are the displayed monthly amounts, not a guarantee of the annual checkout total. (Leonardo pricing)
App plans at a glance
| Plan | Displayed monthly price | Fast-token allowance | Other cost detail |
|---|---|---|---|
| Free | $0 | 150/day | Creations are public; daily allowance resets |
| Essential | $12 | 8,500/month | Private creations and token top-ups |
| Premium | $30 | 25,000/month | Relaxed generation for selected first-party image models |
| Ultimate | $60 | 60,000/month | Relaxed generation for selected first-party image and video models |
Paid plans also differ in model-training allowances and concurrency: the current page lists up to three simultaneous generations on Premium and six on Ultimate. The eligible relaxed models and limits are plan-specific and may change. (Leonardo pricing and plan comparison)
Tokens are a usage meter, not a universal output unit. Leonardo says use depends on the model, feature and computational intensity, including image/video generation and post-processing. An image from one model may cost a different number of tokens from a higher-resolution generation or an upscale. Dividing the plan fee by its token pool therefore does not produce a reliable cost per image unless you also measure the actual settings and token burn.
“Unlimited relaxed” is conditional. It applies only to selected models and runs at lower priority with reduced concurrency. Leonardo's current pricing page identifies eligible models; third-party models can continue to consume tokens. It is not unlimited speed or a promise that every model is included. (Leonardo pricing)
API pricing is separate
Leonardo's API FAQ says PAYG is available now to new API users with no upfront commitment or monthly API fee. Add account balance and requests deduct usage according to the models and features used. An authenticated API pricing calculator estimates an individual request using inputs such as model, resolution and output count. Leonardo does not publish one flat price per image that represents every model and configuration, so this guide does not invent a universal API rate. (Leonardo API pricing FAQ)
Existing API subscription customers are not automatically on the same terms as new PAYG users. Leonardo says their subscription remains unchanged until they choose to migrate. Cancelling the subscription moves the account to PAYG, with eligible remaining credits converted to a non-expiring balance. Verify account-specific migration terms before changing billing. (Leonardo API pricing FAQ)
API credit can be topped up manually or automatically. At zero balance, requests pause until funds are added. Custom per-model discounts are available to eligible customers by contacting Leonardo; they are not the default self-serve rate. (Leonardo API pricing FAQ)
| Workflow | Billing basis | What can be stated safely |
|---|---|---|
| App subscription | Monthly plan plus Fast Token allowance | Price and allowance are published; effective cost per output needs measured token usage |
| Relaxed generation | Lower-priority use for selected models/plans | No extra token charge for eligible first-party models, subject to queue and concurrency rules |
| Production API | PAYG account balance and request usage | Use the logged-in calculator for the exact model/settings; no universal per-image price |
| Existing API subscription | Account-specific transition | Existing terms remain until the customer chooses to migrate, per Leonardo's FAQ |
A useful plan comparison, not fake cost per image
Essential costs $12/month and includes 8,500 Fast Tokens; Premium costs $30/month and includes 25,000. The $18 price difference buys 16,500 more included Fast Tokens. Dividing gives about $0.00109 per additional token ($18 ÷ 16,500). That compares allowance size only; it is not a price per generated image because actions consume different token quantities.
To calculate your real month, record the subscription charge, model and settings per job, tokens used, outputs kept and any top-ups. Then use:
Effective cost per accepted output = (subscription + top-ups) ÷ accepted outputs.
Count rejected generations and editing work. For API use, estimate exact requests with Leonardo's account calculator and keep that balance separate from the consumer app plan.
Which option fits?
- Occasional exploration: try Free if public generations and its daily limit are acceptable.
- Regular private image work: compare Essential's token allowance with a month of measured use.
- High-volume first-party generations: Premium or Ultimate may help if eligible relaxed modes fit your queue tolerance.
- Automated production: use API PAYG when you need programmatic requests, and price exact models/settings in its calculator.
- Mixed app and API work: budget the subscription and API balance separately; do not assume either pays for the other.
For alternatives, see our Nano Banana, Midjourney, FLUX and Ideogram cost comparison, which also distinguishes subscription meters from per-output API pricing.
Bottom line
Leonardo's app gives a known monthly fee and variable token capacity. New API PAYG accounts pay for use without a required recurring API subscription, but the request cost depends on the model and settings. Value depends on what you generate, how many outputs you keep, whether public generations are acceptable and whether relaxed queueing meets your deadline. Start with actual usage, not a guessed number of images per token.
Common questions
Does Leonardo Free include API usage? The app and API are separate products. New API users follow the PAYG terms in Leonardo's API FAQ.
Do paid tokens buy the same number of outputs on every model? No. Usage varies by model, feature and computational intensity.
Can I keep generating after Fast Tokens run out? Selected paid tiers provide relaxed generation for eligible models at lower priority; third-party models may still need tokens.
Are plan prices tax-inclusive? No. Leonardo labels displayed plan prices as excluding tax.
Pricing and terms checked September 26, 2026. Availability may vary by region and account.