Skip to main content

Inference

Per-token Sail API pricing. All inputs include a guaranteed cache write with a 1-hour TTL.

Core models

USDper 1M tokens
ModelTierInputCachedOutput
Kimi K3.1 Pro
Default (ASAP)1.500.2010.00
Balanced1.200.158.00
Flex0.700.106.00
GLM-5.3
Default (ASAP)0.980.183.08
Balanced0.500.122.50
Flex0.400.081.80
GLM-5.3-Flash
Default (ASAP)0.110.020.35
Balanced0.080.020.28
Flex0.050.010.18
DeepSeek V4.1 Flash 1017
Default (ASAP)0.090.0060.30
Balanced0.070.0050.22
Flex0.050.0040.15
DeepSeek V4 Pro 0813
Default (ASAP)0.920.042.77
Balanced0.740.032.22
Flex0.460.021.39
DeepSeek V4 Flash 0731
Default (ASAP)0.090.020.18
Balanced0.070.020.14
Flex0.050.010.09
Gemma 4 31B IT
Default (ASAP)0.400.200.60
Balanced0.120.080.60
Flex0.060.020.30
Gemma 4 31B IT (NVFP4)
Default (ASAP)0.140.070.40
Balanced0.110.060.32
Flex0.070.040.20
Gemma 4 12B IT
Default (ASAP)0.300.152.00
Balanced0.100.072.00
Flex0.050.021.00

Flex-only models

USDper 1M tokens

Models served with the Flex service tier exclusively.

ModelTierInputCachedOutput
Qwen3.6 35B A3B
Flex0.050.020.40

Notes

  • See Service tiers for how to use Balanced and Flex for lower token prices.
    • Not all core models support all tiers yet. We regularly bring up new models and expand service tier support for existing ones based on demand. If you have a need that’s not represented above, get in touch.
  • Prompt caching happens automatically and is guaranteed for all inputs with a TTL of 1 hour.
  • See Models for capabilities and other details on supported models.

Sailbox

Notes

  • Usage accrues only while a Sailbox is running. It is measured every second and added to your bill in one-minute increments.
  • Volume storage is billed separately for each hour the volume exists, even when attached Sailboxes are sleeping, until it’s deleted.
  • Reserved vCPU capacity is shared across your organization’s active Sailboxes. Sleeping and paused Sailboxes release capacity until they resume.
  • For more details, see the Sailbox billing overview.

Pricing Plans

Free

Free

to start, pay-as-you-go

For getting started with Sail

  • $5/month in free credits when you attach a payment method
  • Usage-based pricing with prepaid credits
  • Up to 4 seats
  • 200 reserved vCPUs across active Sailboxes
  • Support via email and the Sail Research Community Slack

Enterprise

Custom

Enterprise-grade customization and support

  • Volume pricing, billed monthly in arrears
  • HIPAA support with a signed BAA
  • Signed MSA and DPA
  • US-only inference with a 10% surcharge on per-token pricing
  • Full usage history through the API, subject to retention
  • Custom model bringups
  • Uptime and latency SLAs
  • Early access to new features
  • Dedicated support via a private Slack channel
  • All additional Pro benefits

Notes

  • You can upgrade to Pro at any time from your billing page.
  • Get in touch to inquire about an Enterprise plan that suits your needs.