More value for your money
We're focused on bringing the CopperRiver experience to as many people as possible — with plans giving much more value than what you pay for.
Free
Try CopperRiver with no commitment
Starter
For individuals getting started with AI
Pro
For professionals who need more power
Requests per model, per plan
Approximate requests per 5 hours. Subject to overall usage limits.
| Model | $9 plan | $19 plan | $49 plan |
|---|---|---|---|
Kimi K2.7 Balanced reasoning | 200 | 460 | 1,100 |
GLM 5.2 Frontier-class, general purpose | 800 | 1,700 | 4,400 |
Minimax M3 Fast and capable | 2,700 | 6,000 | 14,800 |
Qwen 3.7 Plus Coding & tool-use specialist | 3,800 | 8,000 | 21,000 |
Hy3 Highest volume, daily-driver | 5,400 | 11,000 | 30,000 |
Every model is included on every plan. Limits reset every 5 hours.
Common questions
What models are included?
Every plan includes the full lineup: Kimi K2.7, GLM 5.2, Minimax M3, Qwen 3.7 Plus, and Hy3. Switch between them per request — use frontier models for hard problems, fast models for quick queries.
What about my data and privacy?
CopperRiver is designed for international users. Our model providers follow a strict zero-retention policy — your conversations are never stored on their servers and are never used for model training. What you ask stays with you.
Can I cancel anytime?
Yes. Cancel from your account at any time. You keep access until the end of your current billing period — no lock-in, no questions, no retention tricks.
Can I switch between plans?
Absolutely. Upgrade or downgrade whenever you want. Changes are prorated automatically, so you only ever pay for what you actually use.
What about weekly and monthly limits?
Weekly limits are roughly 2x the 5-hour limits shown above, and monthly limits are roughly 4x. So if you pace yourself, you get steady access all month long — limits reset every 5 hours.
How are request limits defined?
Limits are defined in dollar value, not a fixed request count. Your actual number of requests depends on the model you use and how many tokens are in each request — including prompt tokens, cached input tokens, and output tokens. The numbers shown above are based on typical usage patterns. Expensive models yield fewer requests; efficient models yield more.
Who is responsible for what the model does on my machine?
You are. CopperRiver gives models real capabilities — browsing, running terminal commands, reading and writing files, automating tasks. Any action the model takes on your machine, intentional or unintentional, is solely your responsibility. By using the product you agree to accept full responsibility for what the model does. We are not liable for any actions, outcomes, or damages, direct or indirect.