[ Pricing ]
Clear rates. No seat tax.
Start on Quark for free. Pay only for tokens when you ship Origin and NextGen — with automatic volume discounts and cached prompt pricing.
- Billed unit
- Tokens
- Minimum charge
- None
- Invoice cadence
- Monthly
- API keys
- In development
Input and output metered separately per model.
No seat fees on Pro. Pay only for what you use.
Usage invoices close on the 1st. Enterprise can net-30.
Quark is live now. Programmatic keys ship next.
[ Plans ]
Three paths. One meter.
Quark for exploration. Pro for production tokens. Enterprise when you need capacity, networking, and a named channel.
Developer
Developer
Free tier
For exploration and prototyping.
- Origin model access
- 100K tokens / month
- Community support
- Standard rate limits
Pro
Pro
Pay as you go
For production applications at scale.
- All models including NextGen
- Volume pricing
- Priority support
- Higher rate limits
- Usage analytics
Enterprise
Enterprise
Custom
For organizations with advanced requirements.
- Dedicated capacity
- Custom SLAs
- VPC deployment
- SSO & audit logs
- Dedicated support
[ Model rates ]
List prices per million tokens
Input and output are metered separately. Cached identical prefixes bill at the cached rate.
Input
$0.20
/ 1M tokens
Fresh prompts & context
Output
$0.60
/ 1M tokens
Completions & tool calls
Cached input
$0.05
/ 1M tokens
~25% of list input
Volume discounts
List
0 – 10M tokens / mo
List
Pay-as-you-go list rates
Growth
10M – 100M tokens / mo
−10%
Automatic 10% off list
Scale
100M+ tokens / mo
−20%
Automatic 20% off list
Enterprise
Committed capacity
Custom
Custom rates + reserved throughput
[ Calculator ]
See the bill before you ship
Drag tokens, toggle cache share, apply volume tiers. Numbers update live.
[ Cost estimator ]
Volume tier
Estimated monthly
$6.10
Origin · 25M tokens · list rates
- Fresh input
- $2.80
- Cached input
- $0.30
- Output
- $3.00
- List rates
- $0.2 / $0.6 per 1M
Estimates only — actual bills use exact tokenizer counts. Origin details
[ Worked examples ]
What real workloads cost
Ballpark monthly spend for common patterns — not a quote.
[ Origin ]
Support chatbot
10K conversations / day · ~800 input + 200 output tokens each
Input / mo
240M
Output / mo
60M
Est. cost
$80
Origin stays under a coffee budget at this volume.
Rate limits
| Tier | Requests / min | Tokens / min | Monthly |
|---|---|---|---|
| Developer | 60 | 100K | 100K included |
| Pro | 1,000 | 2M | Pay as you go |
| Enterprise | Custom | Custom | Committed + burst |
Soft limits return HTTP 429 before hard spend caps when budget alerts are enabled. Enterprise burst is negotiated.
[ Compare ]
Feature matrix
Focus a column to scan faster. Pro is the default production path.
| Feature | Developer | Pro | Enterprise |
|---|---|---|---|
| Access | |||
| Quark web assistant | ✓ | ✓ | ✓ |
| API access (keys shipping) | Limited | ✓ | ✓ |
| Origin model | ✓ | ✓ | ✓ |
| NextGen model | — | ✓ | ✓ |
| Quark smart routing | ✓ | ✓ | ✓ |
| Limits & cost | |||
| Monthly included tokens | 100K | Pay as you go | Committed |
| Volume discounts | — | ✓ | ✓ |
| Spend caps & alerts | Basic | ✓ | ✓ |
| Cached input pricing | — | ✓ | ✓ |
| Ops & trust | |||
| Usage analytics | Basic | ✓ | ✓ |
| SSO / SAML | — | — | ✓ |
| Audit logs | — | 7 days | Custom retention |
| VPC / private networking | — | — | ✓ |
| Custom SLA | — | 99.9% | 99.95%+ |
| Dedicated support | — | Priority email | Named channel |
[ FAQ ]
Billing questions, answered
[ Next step ]
Use Quark today. Wire the API when keys ship.
Need committed capacity, VPC, or a procurement packet? Talk to enterprise — otherwise open Quark and start building.