Best place to start
$10
Validate a paid model in a real feature without committing to a subscription.
1,777
estimated calls at this workload
Add $10 credits
Prepaid Quan API credits
Powerful financial AI. Start with $10.
Get 100,000 free Quan 3.4 L tokens, then add only the balance you need. No subscription. Credits do not expire. Paid calls stop when your wallet reaches zero.
- 100K free Quan 3.4 L tokens
- No card for the free allowance
- Manual top-ups only
Price a real Quan call
Turn token rates into a workload you can picture.
Adjust the model, token counts, and live web searches. The purchase options update instantly, so you can fund the work instead of guessing at a monthly plan.
01 / Workload
Describe one typical request
Each paid Tavily search adds $0.015. Free Quan 3.4 L usage can include up to two searches while its allowance remains.
02 / Estimate
Your projected cost
Estimated cost per call
$0.0056
$0.0056 tokens $0.0000 grounding
Selected model
Quan 3.4 L
Billing mode
Prepaid balance
Planning estimate, not a quote. Billable output includes visible completion and provider-reported thinking tokens.
03 / Choose a balance
Start small. Keep the ceiling visible.
Every pack funds the same developer wallet and works across public Quan models.
More room to iterate
$25
Compare prompts and model depth while keeping a fixed, predictable spending limit.
4,444
estimated calls at this workload
Add $25 credits
Production buffer
$50
Fund a small production workload with enough room for continued testing and refinement.
8,888
estimated calls at this workload
Add $50 credits
Credit purchases use Stripe Checkout. Email verification and acceptance of the current Developer API Terms are required.
Stripe checkoutSecure one-time payment
No expirationUse the balance when needed
No subscription$0 self-service monthly fee
Hard spending stopPaid calls pause at zero
Transparent model rates
Use more reasoning only when the job earns it.
Prototype with lower-cost models, move high-value analysis to Quan 3.4, and reserve Deep Research for longer multi-stage work.
| Model | Input / 1M | Output / 1M | Best fit |
|---|---|---|---|
| Quan Litequan-3.0 | $0.50 | $3.00 | Fast briefs, routing, extraction, and testing |
| Quan 3.4 Lquan-3.4-l100K free | $0.75 | $4.50 | Cost-aware grounded analysis and repeated finance workflows |
| Quan 3.4quan-3.4 | $2.50 | $15.00 | Flagship analysis, comparisons, filings, and catalysts |
| Deep Researchquan-3.4-deep-research | $4.00 | $24.00 | Longer, multi-step research and portfolio workflows |
Tavily live web grounding is opt-in and passed through at $0.015 per executed search. Free Quan 3.4 L requests can include up to two searches while that allowance remains.
Billing FAQ
The details, plainly.
Prepaid pricing is designed to make the spending boundary obvious before a request runs.
What do I get for free?
A verified developer account receives 100,000 Quan 3.4 L tokens and a separate 100,000-token Quan 3.0 allowance. Free Quan 3.4 L requests can include up to two live searches. Allowances are per account, not per key.
Does the Developer API require a subscription?
No. The self-service Developer API has no monthly fee. Add a prepaid balance only when you need usage beyond an included allowance or want to use a paid model.
Do prepaid credits expire?
No. Developer API wallet funds do not expire, but they belong to the purchasing account and cannot be transferred to another account.
How does the estimator calculate a call?
It applies the selected model’s published input and output rates to your chosen token counts, then adds $0.015 for each paid live web search. It is a planning estimate, not a billing guarantee.
Why might a request cost more than the visible response?
Provider-reported thinking tokens count as billable output. Deep Research can also include internal stages, and paid searches are metered separately.
What happens when my wallet reaches zero?
StockUp reserves estimated cost before paid provider work begins. If the wallet cannot cover the next request, the paid call stops instead of creating an open-ended invoice.
Start with $10. Scale only when Quan earns it.
Fund one real workflow, measure the result, and add more only when the economics make sense.