Pricing
Flexible pricing for your needs.
Packages
Token packages.
Prepay a volume and get a lower per-token price. An account holds one package: every later purchase tops up its limit, and each key chooses whether to spend the package or the balance.
Any model on the price list — 30 today
Solo developers and side projects
Any model on the price list — 30 today
Teams and live products
Any model on the price list — 30 today
Agencies and high-traffic applications
This volume tariff by tariffAny volume on the scale is bought as it stands: the price falls as the volume grows, and a lower rate applies to the tokens past each boundary rather than to the whole purchase — which is what the rows above break down. Beside the price is the same volume at the Pay-as-go rate. Every amount comes from the published price list, to the last digit; per-model charges are visible in the console.
Need more?
Custom volumes, flexible terms and models added on request — tell us what you're building and we'll put a package together.
How it works
From top-up to first call.
Fund a balance
Top up with card or SBP — 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 Kumo. 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.
- googlefrom balance
- openaifrom balance
- anthropicfrom balance
Model rates
Every rate the version publishes.
What a balance pays for a million units of every model, straight off the price list in force.
| About the model | Kumo price | |||||
|---|---|---|---|---|---|---|
| Model | Context | Max output | Capabilities | Input | Cache | Output |
| anthropic | ||||||
| Claude Opus 4.8 | 1M | 128K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 15 USD≈ 1244 ₽ | |
| Claude Sonnet 5 | 1M | 128K | 1.2 USD≈ 100 ₽ | 0.1 USD≈ 10 ₽ | 6 USD≈ 498 ₽ | |
| Claude Sonnet 4.6 | 1M | 128K | 1.8 USD≈ 149 ₽ | 0.2 USD≈ 15 ₽ | 9 USD≈ 746 ₽ | |
| Claude Sonnet 4.5 | 1M | 64K | 1.8 USD≈ 149 ₽ | 0.2 USD≈ 15 ₽ | 9 USD≈ 746 ₽ | |
| Claude Opus 5 | 1M | 128K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 15 USD≈ 1244 ₽ | |
| Claude Opus 4.7 | 1M | 128K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 15 USD≈ 1244 ₽ | |
| Claude Opus 4.6 | 1M | 128K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 15 USD≈ 1244 ₽ | |
| Claude Opus 4.5 | 200K | 64K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 15 USD≈ 1244 ₽ | |
| Claude Haiku 4.5 | 200K | 64K | 0.6 USD≈ 50 ₽ | 0.06 USD≈ 5 ₽ | 3 USD≈ 249 ₽ | |
| Claude Fable 5.1 | 1M | 128K | 6 USD≈ 498 ₽ | 0.2 USD≈ 12 ₽ | 30 USD≈ 2488 ₽ | |
| Claude Fable 5 | 1M | 128K | 6 USD≈ 498 ₽ | 0.6 USD≈ 50 ₽ | 30 USD≈ 2488 ₽ | |
| Claude Opus 5.5 | 1M | 128K | 2.2 USD≈ 182 ₽ | 0.1 USD≈ 9 ₽ | 11 USD≈ 912 ₽ | |
| openai | ||||||
| GPT-5.6 Terra | 1.1M | 128K | 1.2 USD≈ 100 ₽ | 0.1 USD≈ 10 ₽ | 7.2 USD≈ 597 ₽ | |
| GPT-5.6 Sol | 1.1M | 128K | 2.4 USD≈ 199 ₽ | 0.2 USD≈ 20 ₽ | 12 USD≈ 995 ₽ | |
| GPT-5.6 Luna | 1.1M | 128K | 0.1 USD≈ 10 ₽ | 0.01 USD≈ 1 ₽ | 0.7 USD≈ 60 ₽ | |
| GPT-5.5 | 1.1M | 128K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 18 USD≈ 1493 ₽ | |
| GPT-5.3 Codex Spark | 128K | 32K | — | 1.1 USD≈ 87 ₽ | 0.1 USD≈ 9 ₽ | 8.4 USD≈ 697 ₽ |
| Codex Auto Review | 1.1M | 128K | 3 USD≈ 249 ₽ | 0.3 USD≈ 25 ₽ | 18 USD≈ 1493 ₽ | |
| GPT-6 Astra | 1.1M | 128K | 6 USD≈ 498 ₽ | 0.6 USD≈ 50 ₽ | 30 USD≈ 2488 ₽ | |
| GPT-6 Sol | 1.1M | 128K | 1.1 USD≈ 91 ₽ | 0.1 USD≈ 9 ₽ | 5.5 USD≈ 456 ₽ | |
| Gemini 3.8 Flash High | 1M | 65.5K | — | 0.5 USD≈ 37 ₽ | 0.05 USD≈ 4 ₽ | 2.3 USD≈ 187 ₽ |
| Gemini 3.7 Flash High | 1M | 65.5K | — | 0.5 USD≈ 37 ₽ | 0.05 USD≈ 4 ₽ | 2.3 USD≈ 187 ₽ |
| Gemini 3.6 Flash High | 1M | 65.5K | — | 0.5 USD≈ 37 ₽ | 0.05 USD≈ 4 ₽ | 2.3 USD≈ 187 ₽ |
| Gemini 3.1 Flash Lite | 1M | 65.5K | 0.2 USD≈ 12 ₽ | 0.02 USD≈ 1 ₽ | 0.9 USD≈ 75 ₽ | |
| Gemini 3 Flash Preview | 1M | 65.5K | — | 0.3 USD≈ 25 ₽ | 0.03 USD≈ 2 ₽ | 1.8 USD≈ 149 ₽ |
| xai | ||||||
| Grok 4.6 | 500K | 500K | 1.2 USD≈ 100 ₽ | 0.3 USD≈ 25 ₽ | 3.6 USD≈ 299 ₽ | |
| Grok 4.5 | 500K | 500K | 1.2 USD≈ 100 ₽ | 0.2 USD≈ 15 ₽ | 3.6 USD≈ 299 ₽ | |
| Grok 4.7 | 500K | 500K | 1.1 USD≈ 91 ₽ | 0.3 USD≈ 23 ₽ | 3.3 USD≈ 274 ₽ | |
| zai | ||||||
| GLM-5.3 | 1M | 131.1K | 0.8 USD≈ 70 ₽ | 0.2 USD≈ 13 ₽ | 2.6 USD≈ 219 ₽ | |
| GLM-5.2 | 1M | 131.1K | 0.8 USD≈ 70 ₽ | 0.2 USD≈ 13 ₽ | 2.6 USD≈ 219 ₽ | |
Balance rates are quoted per million units and rounded to one decimal; the published rate appears on hover. A dash means the version in force publishes no rule.
Security & trust
Built for production, not just demos.
Security and reliability are built into Kumo from day one. Three commitments — each backed by an article you can read in full.
Your data isn't trained on
Prompts and completions are never used to train models. Content is processed only to serve your request, then dropped — we retain just the account and usage metadata needed to operate the service.
Read the trust centerNo prompt-body retention
We don't log or store the body of your prompts — only usage metadata for billing. What we keep, for how long, and how it's processed is written down in the privacy policy and DPA.
Privacy & retentionTransparent logic, 99.9% target
One gateway, no silent model swaps: you get the exact model you asked for, upstream errors surface as a contract-shaped 502, and we run to a 99.9% uptime target.
How it worksFAQ
Pricing, answered.
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.