Pricing

Pay per token.
Top up in Rupiah.

Three models at fixed, published rates. Add balance with QRIS, and each request is charged for the tokens it actually used. No subscription, no monthly minimum, no expiry.

DeepSeek-V4.1-Flash

USD per 1M tokens

deepseek-flash
Input · cache hitPrompt tokens the provider served from its cache
$0.010
Input · cache missPrompt tokens processed fresh
$0.400
OutputCompletion tokens the model generates
$1.600
Context window
1M tokens
Maximum output
384K tokens
Default output cap
8,192 tokens
Price by time of day
Fixed

DeepSeek-V4-Pro

USD per 1M tokens

deepseek-v4-pro
Input · cache hitPrompt tokens the provider served from its cache
$0.060
Input · cache missPrompt tokens processed fresh
$1.760
OutputCompletion tokens the model generates
$5.280
Context window
1M tokens
Maximum output
384K tokens
Default output cap
8,192 tokens
Price by time of day
Fixed

GLM-5.3-Flash

USD per 1M tokens

glm-flash
Input · cache hitPrompt tokens the provider served from its cache
$0.040
Input · cache missPrompt tokens processed fresh
$0.200
OutputCompletion tokens the model generates
$0.667
Context window
1M tokens
Maximum output
128K tokens
Default output cap
8,192 tokens
Price by time of day
Fixed

Cost calculator

Estimate what your workload costs.

Enter the token split of a typical request. The math is the same formula the ledger uses, so the result matches what your console would show.

Cost breakdown for one request on DeepSeek-V4.1-Flash
Per requestTokensCost
Cache-hit input × $0.010 / 1M8,000$0.000080
Cache-miss input × $0.400 / 1M4,400$0.001760
Output × $1.600 / 1M1,800$0.002880
One request$0.004720
1,000 requests$4.720000
Requests a $10 top-up covers≈ 2,118

Estimates use list prices. Actual charges use the token counts the provider reports for each request.

Top-ups

What a top-up buys.

Any amount from $1.00 to $100.00, in whole cents. You pay the IDR equivalent at the rate on your invoice. Unused balance stays in your wallet.

Approximate tokens per top-up amount
CreditDeepSeek-V4.1-FlashDeepSeek-V4-ProGLM-5.3-Flash
≈ Cache-miss input≈ Output≈ Cache-miss input≈ Output≈ Cache-miss input≈ Output
$1.002.5M0.63M0.57M0.19M5M1.5M
$10.0025M6.25M5.68M1.89M50M14.99M
$25.0062.5M15.63M14.2M4.73M125M37.48M
$100.00250M62.5M56.82M18.94M500M149.93M

FAQ

Billing questions.

Something not covered? Email support@larasrouter.com.

Why are model prices listed in USD if I pay in Indonesian Rupiah?

Our model costs are billed to us in USD, so usage is priced and held in USD. When you top up, you pay the IDR equivalent via QRIS at our current exchange rate. The full USD amount you chose is credited to your balance.

How is the exchange rate set?

We set the USD-to-IDR rate ourselves and review it weekly. It includes a conversion buffer over the mid-market rate (currently about 2%) to cover currency movement. Each invoice shows the rate and the IDR total before you pay, and that rate is fixed for the invoice’s 15-minute validity.

Does my wallet balance fluctuate when exchange rates change?

No. Once your payment succeeds, your balance is held in USD. Later exchange-rate changes only affect future top-ups.

What are the top-up limits? Are there monthly fees?

Each top-up is between $1.00 and $100.00, in whole cents. There are no monthly fees, inactivity penalties, or seat licenses, and your balance never expires. There is no free or promotional credit, so your first request needs a top-up.

Which Indonesian payment applications are supported for QRIS?

Any QRIS-enabled banking or e-wallet app, for example BCA Mobile, myBCA, Livin’ by Mandiri, BRImo, BNI Mobile, GoPay, OVO, ShopeePay, DANA, or LinkAja.

How is token usage calculated and deducted?

Each request is charged at the rates of the model it calls, using the token counts the model's provider reports. Per 1M tokens, as cache-hit input / cache-miss input / output: DeepSeek-V4.1-Flash $0.010 / $0.400 / $1.600; DeepSeek-V4-Pro $0.060 / $1.760 / $5.280; GLM-5.3-Flash $0.040 / $0.200 / $0.667. Before a request runs, we reserve a hold on your balance; afterwards the hold is settled to the actual charge and the rest is released. Charges are shown to six decimal places ($0.000001) in your console.

Why did a request return 402 insufficient_credit?

Requests run while your available balance is above $0.00, even when it is smaller than a request's hold (about DeepSeek-V4.1-Flash $0.41; DeepSeek-V4-Pro $1.79; GLM-5.3-Flash $0.20 at the default output cap, depending on the model). A request that finishes can take your balance below zero by at most $1.00, and that amount is deducted from your next top-up. While your available balance — after holds for requests still running — is $0.00 or below, new requests return 402 insufficient_credit until you top up or those requests finish. If the error code is key_budget_exceeded instead, the request would push that API key past its monthly budget; raise the budget in the console or wait for the next month.

Am I charged for failed requests?

A request that the provider rejects with a 4xx error is not charged. If a request fails with a 5xx error, times out, or is interrupted, its hold is kept while we reconcile it: you may be charged for any usage the provider reports, otherwise the hold is released without a charge.

Can I withdraw or refund my wallet balance?

No. Top-ups are final prepaid credit purchases: credited funds cannot be withdrawn, refunded, or redeemed for cash. Your balance never expires. If you are charged twice or incorrectly, contact support@larasrouter.com and we will correct it on your balance.