Read a price
Every model and preset has apricing object:
So a 4-second clip from
google/veo-3.1-fast with sound costs 4 × $0.20 = $0.80, and two images from black-forest-labs/flux-2-pro cost 2 × $0.041 = $0.082.
A key in when, like a name in drivers, can be a dotted path into a group of settings. recraft/recraft-v3 costs $0.054 an image, and $0.11 with { "when": { "settings.style": "vector_illustration" } }.
length_of and length_step appear only on prices by the length of your own file. This is openai/whisper-large-v3, which bills each started minute of the recording at $0.0027:
Get a quote
You don’t have to do that sum.POST /v1/quotes takes the same body as a generation and returns its price, with the input checked and its defaults filled in. It doesn’t run anything or hold any credit.
A quote is optional. A script can start the generation straight away: if the balance doesn’t cover it, the request gets a 402 and nothing runs. Use a quote to show a price to a person in your app before they press go, or to check an input without spending.
POST /v1/generations would answer with a 400 gets the same 400 here, for free. It can’t tell whether the model will accept your prompt: a refusal shows up later, as a failed generation that isn’t charged.
A quote doesn’t reserve its price, and a run is priced again when it starts. A stable model’s price changes only after notice; a preview model’s can change at any time. For a model priced by the length of your own file, the final charge is known only when the run ends (see Holds and charges).
Holds and charges
When a run starts, its price is held from your balance. When it ends:- If it succeeded, you’re charged
cost_usd, and the generation’susage.cost_usdshows it. - If it failed, was refused or was canceled, the hold is released and nothing is charged.
hold_usd is usually the same as cost_usd. For a model priced by the length of your own file, it can be more: the price of your file’s length and a second more, or of the longest file the model takes when your file’s exact length isn’t known. Such a run is charged when it ends, at the quoted cost_usd, or more if the model counted a longer length than your file’s.
A run starts when your available balance covers its cost_usd. If the balance covers the price but not the second more, the run holds what you have. If the hold is the longest file the model takes, the balance has to cover the whole hold. Otherwise the request gets a 402 with the code insufficient_credit, and nothing starts.
Your balance
available_usd is what you can spend now: your balance minus the holds of runs still going. held_usd is what those runs hold.
auto_top_up says whether the balance tops itself up: when it falls below threshold_usd, Leap adds amount_usd from the card your last top-up saved. add_credits_url is where a person adds credit and turns it on.
welcome says whether your workspace got the one-time welcome credit. Workspaces made since 14 October 2026 don’t get one and read none: you add credit before you run anything, and your first top-up has no fee. For a workspace made before then, status is granted, with amount_usd granted; withheld, when the sign-up went past the limits that stopped sign-up farms, so none came and none will; or none, when it wasn’t a workspace that got one. A welcome credit doesn’t count as paid credit, so it doesn’t let an API key run (see Before your first top-up).
Add credits, and turn on auto top-up so a job never stops for an empty balance, at app.tryleap.ai/go/settings/credits.
Auto top-up and card fees
A top-up under $20 has a card fee of 5.5%; from $20 there is none, and your workspace’s first top-up has none at any amount. Auto top-up charges the card a top-up saved, so it can be turned on once you’ve topped up after your first one (a first top-up saves no card). Set it to add $20 or more and it never pays a fee. When a key runs out of credit, its402 names the credits page, and says so when auto top-up is off.
Before your first top-up
API keys run on paid credit. Until your workspace’s first top-up, a run or a batch started with a key gets a402 with the code top_up_required, whatever the balance, and nothing starts. The first top-up has no fee, and your keys run as soon as it goes through. If every top-up is later refunded or disputed in full, they stop again until the next one.
In the studio, each new person’s first image is free: one picture from a model page, a recipe or the studio’s first screen, at settings that cost up to $0.25 to make, never a clip or a sound, and once per person and per email inbox. A model or recipe whose first image is free at its defaults says so with first_image_free: true. A run started with an API key or over MCP is never free. Every other run needs credit.
In the studio and over MCP signed in with your Leap account, a workspace that hasn’t added credit yet meets these limits. All of them lift with its first top-up, unless that top-up is refunded.
Once 6 video runs have ended without a charge, video runs wait for the first top-up. Images keep running. A refusal by the model’s safety filter, a blocked photo, a failure after the run’s preview and a cancel after it count; provider errors before a preview, timeouts, cancels before a preview and outages don’t. Only workspaces created since this limit shipped meet it.
One run priced by the length of your own video or sound runs at a time. A second one waits for the first to finish.
Speech-to-text and captions models, such as openai/whisper-large-v3, transcribe up to 120 minutes per UTC day. Each run counts its started minutes, and a run that fails or is canceled gives them back. A request for more than 120 minutes needs credit, and once the day’s minutes are used up, the next run waits for midnight UTC.