Skip to content
Kumo
AI API Gateway

The best models. One endpoint.
30–50% less.

One API for supported GPT, Claude and Gemini models — one key, one balance, prices 30–50% lower. Every model is the original — no swaps, no downgraded versions.

Get your API keyExplore Models

No credit card · one base URL · no VPN · free to start

lower effective spend
30–50%lower effective spend
active text models
9active text models
uptime target
99.9%uptime target

Models supported

GPT-5.5GPT-5.4GPT-5.4 miniClaude Opus 4.8Claude Sonnet 4.6Claude Haiku 4.5Gemini 2.5 ProGemini 2.5 FlashGemini 2.5 Flash-Lite

Models are provided by their labs. Kumo is not affiliated with or endorsed by them.

Packages

Token packages

One balance, one key. Top up and pay per token — no subscriptions, no minimums.

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.

Full pricing & calculator

Models

Recommended models for production.

A curated set of production-ready GPT, Claude and Gemini picks at Kumo's effective rate.

Best value

GPT-5.4

OpenAI
−30%
Input / 1M
$2.50$1.75
Output / 1M
$15.00$10.50

Same model, same answer — 30% lower effective rate.

Get your API key
Top pick

Claude Opus 4.8

Anthropic
−30%
Input / 1M
$5.00$3.50
Output / 1M
$25.00$17.50

Same model, same answer — 30% lower effective rate.

Get your API key
Fastest

Gemini 2.5 Pro

Google
−30%
Input / 1M
$1.25$0.88
Output / 1M
$10.00$7.00

Same model, same answer — 30% lower effective rate.

Get your API key

From signup to first call

Three steps to your first request.

Sign up, top up once, grab a key — and you're calling supported GPT, Claude and Gemini text models from one balance. No sales call, no contract, no per-seat fees.

1

Sign up

Create an account to get started. Add your team to an org whenever you're ready.

2

Top up once

Buy credits once, then spend them on any model or provider from a single balance.

$0
Top up to fund your balance
3

Get your API key

Create a key and start shipping. Fully OpenAI-compatible — change one base URL.

KUMO_API_KEY
••••••••••••••••

Generate a key to see your token budget.

Who it's for

The right token price for every team.

Developers, AI agents, products, enterprise — we shape Kumo around how each one uses tokens: the same models for less, fast and without a VPN, and the difference goes straight back into your margin.

Rates verified Jun 2026

What you get

Same models, same code — your token bill drops 30–50%.

Drop-in via the OpenAI API: switch one base URL, keep the rest of your code — Cursor, Cline, Roo, Continue or your own scripts. You pick every model yourself: low-cost DeepSeek and Qwen, flagship Claude or GPT — call whatever fits the task, billed at Kumo's optimized rate. Pay only for real tokens — no seats, no subscription, no VPN.

Get your API key

Volume / month

List / month
$60
Your price / month
Claude Sonnet 4.6
$42
Savings / month
$18−30% vs list

↳ that's ≈ 5 features (~1K LOC) · ≈ 3 days in Claude Code

Requests / day
152
~3,000 tokens / request
Input / 1M
$3.00
input tokens
Output / 1M
$15.00
output tokens
Blended / 1M
$6.00
at 25% output

Indicative estimate at a 75% input / 25% output split. "List" is the provider's standard rate; "Your price" is Kumo's automatically optimized effective rate. Tokens per unit are conservative estimates. Exact rates are visible in your dashboard before you're charged.

Security & trust

Built for production, not just demos.

Security and reliability are built into Kumo from day one. Here's how we protect your data and keep the service dependable.

No training
your data is never trained on
No prompt-body retention
we don't keep your prompts
99.9%
uptime target

Security & data

Your data isn't trained on

Prompts and completions are never used to train models. Content is processed to serve your request and nothing more.

Prompt content isn't logged

We store usage metadata for billing — not the body of your prompts. Bring your own retention policy.

Prompt bodies are not retained

Kumo forwards prompts and completions to serve the call, then drops the content. We retain only the account, billing and usage metadata needed to operate the service.

Why Kumo

Every provider, one key

OpenAI, Anthropic and Google models behind a single API key. If a provider errors, the gateway returns a 502 and you can switch providers directly.

Token accounting

Real-time spend, per-key limits and detailed logs. Overspend shows up the moment it happens — no surprise invoice at month's end.

Drop-in compatible

Works with Claude Code, Cline, Cursor, Kilo Code, Opencode and any OpenAI SDK — no code changes. Swap one base URL and you're done.

Trust center →Uptime & SLA →Privacy & retention →Data processing (DPA) →Quickstart & API →

How people use Kumo

One balance, supported models — put to work.

From a solo build on nights and weekends to agents running 24/7 and an agency's whole AI stack — here's how different teams actually run on Kumo.

Next steps

See how it works — or talk to us.

The full picture of how Kumo runs under the hood, plus a direct line for volume deals and consulting.

How it works

One gateway in front of supported models.

Point one base URL at Kumo and we handle routing, billing and team access — then hand the optimized rate back to you.

  • The request lifecycle, call by call
  • Real lab accounts — no reseller chain
  • Why it's 30–50% cheaper, with the mechanism
  • Drop-in migration in one line
See how it works
Direct contacts

Talk to the team.

We reply within one business day.

B2B & consulting

Moving serious volume?

Volume discounts, company invoicing and pay-by-contract — plus hands-on help embedding Kumo into your stack.

Submit a request
Developers

API documentation.

OpenAI-compatible endpoints, SDK snippets and a quickstart — everything to make your first call.

Read the docs
Quick message

Tell us what you need.

A line about your project or question — we'll route it to the right person.

Write a message

Make your first call tonight.

Get your API key — free→Read the API docsContact us

Moving serious volume? Volume discounts, company invoicing and hands-on consulting — talk to us