Pricing & Credits

AI Playground uses pay-per-generation credits — no per-provider subscriptions, no API keys to manage. Every call shows its credit cost before you commit, and one balance covers the available models.

Check your balance

gen-ai credits

{ "name": "picsart_credits", "arguments": {} }

Quote a cost (dry run)

Always free — no model is invoked and nothing is charged. gen-ai pricing needs a signed-in session, because prices are looked up for your account.

gen-ai pricing seedance-2.0 --duration 5 --resolution 1080p   # cost for a specific config
gen-ai pricing veo-3.1 --duration 8 --audio                   # 8-second Veo clip with audio
gen-ai pricing --mode video --json                            # all video model pricing

gen-ai pricing narrows by --duration, --resolution, and --[no-]audio; other parameters are not part of the CLI quote.

{ "name": "picsart_preflight",
  "arguments": {
    "model": "veo-3.1",
    "params": { "prompt": "a drone shot over a snowy ridge", "duration": 8, "resolution": "1080p" }
  } }

picsart_preflight validates the params and quotes the cost in one free call, returning { model, valid, errors?, credits }. credits is a number, or null if pricing isn't available for that model or the call is unauthenticated. For the balance rather than a per-call cost, use picsart_credits.

What drives cost

Cost depends on the model and its parameters. The biggest factors:

• Video — duration, resolution, and whether audio is generated.
• Image — resolution/quality and the number of outputs (count).
• Audio — length of the output.

Because pricing is resolved per-model and per-params, the only reliable number is the one from a live pricing quote against the exact payload you intend to run.

Compare before you commit

Quote the same prompt across candidates to find the best value:

gen-ai pricing seedance-2.5 --duration 8
gen-ai pricing veo-3.1 --duration 8
gen-ai pricing seedance-2.0 --duration 8

FAQ

Is gen-ai pricing always accurate?

It reflects the current cost for the exact model and parameters you pass. Cost can change if the model's pricing is updated, so always run a fresh quote before a large batch run.

Why does picsart_preflight return null for some models?

A few models do not expose per-call pricing. For those, picsart_preflight returns null for the credit amount. Check the model's page in the Model Reference for any fixed or range-based pricing notes.

Are there per-provider contracts or subscriptions?

No. One Picsart account covers all 29 providers. There are no separate subscriptions, no per-provider API keys, and no vendor invoices.

What happens if I run out of credits mid-batch?

Completed generations in the same batch are not reversed. Use gen-ai batch resume <output-dir> (e.g. ./batch-output) to re-run the failed jobs after topping up.

Can I set a spending limit per run?

Per generation, yes: gen-ai generate … --max-cost <credits> aborts before submitting if the estimated cost is higher (the same flag works on the other generation commands such as image and video). There is no total cap for a batch run — quote representative items with gen-ai pricing or picsart_preflight first and estimate the total from there.