Free
$0
no card required
Test every model. Free credits to evaluate.
- $5 in starter credits
- All public models
- Up to 60 RPM, 60k TPM
- Community support
Pricing
Per token. No idle compute. Caps you can trust.
You pay for tokens used, not GPUs sitting idle. Cached input tokens bill at 50% off automatically. Fine-tuned variants price the same as the base model.
Pro
most popularPay as you go
per-token billing
Production traffic, transparent telemetry, hard spend caps.
- All models, all features
- Up to 600 RPM, 600k TPM
- Hard spend caps & alerts
- P99 latency dashboard
- 50% off cached input tokens
- Email support
Enterprise
Custom
annual contract
VPC isolation, single-tenant clusters, contractual SLAs.
- Everything in Pro
- Single-tenant on Trainium / GPU
- P99 latency SLA
- SOC 2 Type II (via Decart AI)
- GDPR + EU data residency
- HIPAA BAA on roadmap
- VPC peering, zero egress
- Dedicated solutions engineer
Calculator
Math we don't hide.
Plug in monthly volume to see what your bill actually looks like.
Model
Input tokens2.0M / month
Output tokens1.0M / month
Cached input share0% of input tokens repeat (50% discount)
All models, all prices
Updated quarterly. Past pricing posted in the changelog.
| Model | $/1M input | $/1M output | Cached input |
|---|---|---|---|
GPT-OSS 120B OpenAI | $0.04 | $0.18 | $0.02 |
DeepSeek V4 Pro DeepSeek | $0.43 | $0.87 | $0.22 |
DeepSeek V4 Flash DeepSeek | $0.14 | $0.28 | $0.07 |
Kimi K2.6 Moonshot | $0.74 | $3.49 | $0.37 |
DeepSeek V3.2 DeepSeek | $0.25 | $0.38 | $0.13 |