Start small. Scale when the traffic arrives.
Choose a credit pack, then pay only for the tokens you use.
Starter
$20
A focused pack for prototypes and evaluation.
- ✓Prepaid inference credits
- ✓Virtual API keys
- ✓Usage dashboard
Growth
$50
More headroom for active products and teams.
- ✓Everything in Starter
- ✓Automatic failover
- ✓Streaming completions
Scale
$100
A practical starting point for production traffic.
- ✓Everything in Growth
- ✓Higher usage capacity
- ✓Design-partner eligibility
Model rates
Credits are spent per model at these rates. Input served from the provider's prompt cache is billed at 10% of the input rate.
| Model | Input / 1M tokens | Cached input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
| Kimi K3kimi-k3 | $2.50 Save 17% | $0.25 | $12.50 Save 17% |
| DeepSeek V4 Pro 0813deepseek-v4-pro-0813 | $0.435 | $0.0435 | $0.87 |
| DeepSeek V4 Flash 0731deepseek-v4-flash-0731 | $0.11 Save 21% | $0.011 | $0.25 Save 11% |
Savings compare Corion rates with the provider's official list prices (Moonshot AI, DeepSeek), as of August 2026. Official prices may change.
Design Partner
A small group of partners shaping our capacity, latency, and pricing, with terms to match. If frontier open models are core to your product, we want to hear from you.
Talk to us →Enterprise
Priority tiers with dedicated SLA, priority capacity, and US-only serving options. On-demand deployment and fine-tuning available on inquiry.
Contact sales →