Pricing

Two ways to pay. One lower bill.

Pay only for the tokens you use from a pay-as-you-go balance, or commit to a volume up front and buy it as a package at a deeper rate. Either way it's the same GPT, Claude and Gemini — billed 30–50% below list.

Prices follow the currency switch in the header (USD / RUB).

Choose how you pay

Balance or package.

Both draw from one prepaid balance behind one OpenAI-compatible key. The difference is only when you commit — and how deep the rate goes.

Pay-as-you-go

Top up your balance

Flexible. Pay per token, any model, no commitment.

Top up once from $5 — spend it on any model from one balance.
Every call draws down at that model's rate, 30–50% below list.
No plans, no seats, no minimums. Auto-recharge optional.
Unused balance remains spendable while your account stays active; token purchases are final.

Top up & get a key

Prepaid · bulk

Buy a token package

Commit volume up front — earn a deeper committed-use rate.

Pick a model and a committed volume — pay once, lock the rate.
The bigger the package, the deeper the rate: up to −20% on top of the base saving.
Or take a multi-model bundle — one balance across a curated set.
Same tokens, same models — just bought at the committed-volume price.

Build a package

Packages

Token packages.

The full catalog — prepay a volume and get a lower per-token price. Every package funds the same single balance behind your key.

ПОПУЛЯРНЫЙ

60,000,000 tokens

≈20,000 API requests on a single balance

$0.34 / 1M tokens

$20.62

≈ 1 639 ₽

Don't need a package?Just top up a balance and pay as you go

Pay in rubles — by card or SBP; companies get an invoice with closing documents.

Indie developers

Side projects and experiments

API access

3,000,000 tokens

$0.75 / 1M tokens

~1,000API requests
~1day of heavy API use

$2.25

≈ 179 ₽

API access

5,000,000 tokens

$0.68 / 1M tokens−10% per token vs entry

~1,667API requests
~1.7days of heavy API use

$3.38

≈ 269 ₽

API access

10,000,000 tokens

$0.50 / 1M tokens−33% per token vs entry

~3,333API requests
~3.3days of heavy API use

$5.02

≈ 399 ₽

Products

A key inside your product — lower token COGS

API access

15,000,000 tokens

$0.45 / 1M tokens

~5,000API requests
~5days of heavy API use

$6.78

≈ 539 ₽

API access

20,000,000 tokens

$0.44 / 1M tokens−3% per token vs entry

~6,667API requests
~6.7days of heavy API use

$8.79

≈ 699 ₽

API access

35,000,000 tokens

$0.40 / 1M tokens−13% per token vs entry

~11,667API requests
~12days of heavy API use

$13.83

≈ 1 099 ₽

API access

50,000,000 tokens

$0.36 / 1M tokens−20% per token vs entry

~16,667API requests
~17days of heavy API use

$18.10

≈ 1 439 ₽

Companies

Pooled team volume, one invoice

API access

60,000,000 tokens

$0.34 / 1M tokens

~20,000API requests
~20days of heavy API use

$20.62

≈ 1 639 ₽

API access

80,000,000 tokens

$0.31 / 1M tokens−9% per token vs entry

~26,667API requests
~27days of heavy API use

$25.15

≈ 1 999 ₽

API access

100,000,000 tokens

$0.30 / 1M tokens−12% per token vs entry

~33,333API requests
~33days of heavy API use

$30.18

≈ 2 399 ₽

API access

150,000,000 tokens

$0.26 / 1M tokens−24% per token vs entry

~50,000API requests
~50days of heavy API use

$38.99

≈ 3 099 ₽

API access

200,000,000 tokens

Best value

$0.24 / 1M tokens−30% per token vs entry

~66,667API requests
~67days of heavy API use

$47.80

≈ 3 799 ₽

Calculator

Estimate your price.

Pay-as-you-go or a prepaid package — pick a model, size it, and see your effective Kumo rate against list. Same models, optimizations on by default.

Rates verified Jun 2026

Model

Input tokens / month20M

20M

Output tokens / month5M

$5 in · $25 out · per 1M tokens

You save

$68/mo

≈ $810 a year · 30% off list

List

$225

On Kumo

$158

Save $68/mo — get your key

Same model, same API · pay only for what you use. Indicative estimate from list-price deltas; exact rates show in your dashboard before you spend.

Multi-model bundles

One package, a whole stack.

Prefer a ready set? Each bundle is one prepaid balance spendable across a curated group of models — bought at the bundle's committed-volume rate.

Starter

Fast, cheap models to prototype on.

$10$13

−22% off list · $1 / 1M

Prepaid volume: 10M
Models: Claude Haiku 4.5 · Gemini 2.5 Flash-Lite · GPT-5.4 mini

Side projects & first calls

Get this bundle

Builder

Scale

Flagship models at the deepest rate.

$150$6,354

−98% off list · $0.15 / 1M

Prepaid volume: 1B
Models: Claude Opus 4.8 · Gemini 2.5 Pro · GPT-5.4

Agencies & high-volume apps

Get this bundle

Bundle prices are estimates at the average blended rate across the included models (25% output split). You spend the balance on any model in the set; exact rates show in your dashboard.

Bulk discount

Buy more, pay less per token.

Committing to volume up front earns a deeper rate — passed straight back to you on top of the model's standard saving. The bigger the commitment, the bigger the bonus.

Committed volume	Bulk bonus	Total off list*
10M	— (base saving)	−30%
50M	+5%	−35%
100M	+10%	−40%
500M	+15%	−45%
1B	+20%	−50%

* Example total on a model with a 30% base saving (e.g. Claude Sonnet 4.5). The bulk bonus stacks on top of whatever the model already saves.

Funding

Fund it your way.

Both the balance and packages take fiat or crypto. Pick whatever's easiest — the price is the same.

Card & bank transfer

Pay by card or bank transfer in USD. Funds land on your balance instantly and you're calling models within the minute.

Crypto

Top up with USDT, BTC or ETH. The balance is credited once the transfer confirms on-chain — no card required.

One balance, either way

However you fund it, it's one prepaid balance behind one key — spend it on any model, top up again whenever.

How it works

From top-up to first call.

Fund a balance

Top up with card, bank transfer or crypto — pay-as-you-go, or buy a committed package for the deeper rate.

Grab one key

Create an OpenAI-compatible key and point your base URL at api.kumo.cloud/v1. Your code learns nothing new.

Spend across supported models

Every call draws from the same balance at the rate you saw. Watch usage and remaining balance live in the dashboard.

FAQ

Pricing, answered.

Start with a balance: top up, pay per token, no commitment — ideal while your volume is unpredictable. Once you know you'll spend a known volume on a model, buy it as a package up front and the committed-volume bonus lowers the per-token rate. You can do both; packages and balance top-ups land in the same wallet.

Balance or package — which should I pick?

How does the bulk discount actually work?

Simple volume economics, passed back to you: the more you commit to up front, the deeper the bonus on top of the model's standard rate — up to −20% extra at the 1B-token tier. It's not a limited-time sale; it's a straightforward price break for buying in bulk, the same way it works anywhere else.

Do packages cost the same in crypto and fiat?

Yes. Fiat (card, bank transfer) and crypto (USDT, BTC, ETH) are just ways to fund the same prepaid balance — the package price and the per-token rate are identical either way.

Do prepaid tokens expire?

Unused balance remains spendable while your account stays active. Token purchases are final and non-refundable; the rate you locked applies to the package volume.

What's the output/input split in these estimates?

The figures on this page assume a 25% output share — typical mixed traffic. Your real cost depends on your own input/output mix; the dashboard shows the exact per-model input and output rates before you spend.

Can I still switch models on a package?

A single-model package is priced for that model. Need several models? Take a multi-model bundle — one prepaid balance spendable across the supported set — or just use a pay-as-you-go balance, which works across supported models with no commitment.

Top up once. Call supported models.

Start on a pay-as-you-go balance for free, or commit to a package for the deeper rate — your call.

Get your API key Contact us