Pricing Plans

Choose a monthly plan.

Get 5,000 monthly model transactions for free. Get a free API key.

Deploy models to a single data center with automatic failover.

Personal
$9
  • 75,000 Monthly Predictions
  • No Latency Routing (?)
  • -
  • 1 Model
  • Scikit-Learn Models up to 100 MB
  • Unlimited Online Model Training
  • Email Support
Get Started
Light
$15
  • 300,000 Monthly Predictions
  • No Latency Routing (?)
  • -
  • 2 Models
  • Scikit-Learn Models up to 100 MB
  • Unlimited Online Model Training
  • Email Support
Get Started
Startup
$25
  • 1.25 Million Monthly Predictions
  • No Latency Routing (?)
  • Burst Protection (?)
  • 3 Models
  • Scikit-Learn Models up to 100 MB
  • Unlimited Online Model Training
  • Email Support
Get Started
Business
$49
  • 3 Million Monthly Predictions
  • No Latency Routing (?)
  • Burst Protection (?)
  • 5 Models
  • Scikit-Learn Models up to 100 MB
  • Unlimited Online Model Training
  • Email and Phone Support
Get Started

Contact us for additional pricing options if you need more monthly predictions, more models, or an SLA.

Frequently Asked Questions

Email support@mlrequest.com if you have other questions.

  • What payment methods do you accept?

    We accept all major credit and debit cards.

  • What is the difference between single and multi-region plans?

    Single-region plans will deploy models to a single data center of your choosing and multi-region plans will deploy models to all 5 data centers around the world. Both plans will automatically failover when model servers become unavailable. Currently the data centers are located in the Eastern United States, Western United states, Germany, India, and South Korea.

    A single-region plan works well when prediction latency isn't important, or the users of your machine learning-based application are centrally located (e.g., all in the United States). Response times will completely depend on the client location and can be as little as 30ms or as much as 1000ms.

    Choose a multi-region plan to take advantage of latency-based routing, which will achieve < 60ms response time on average from anywhere on earth. This plan type is best for applications with globally distributed users and when response time is important.

  • What is burst protection?

    Burst protection provides you with extra transactions when you are past your monthly limit. Extra transactions will be provided for up to 5 days, at which point we will ask you to upgrade to another plan.

  • What is latency routing?

    Latency-based routing provides the fastest response times (e.g., for prediction requests), by routing your requests to the region with the lowest latency. On average, latency-based routing achieves response times of < 60ms.

    Latency-based routing is enabled for plans that deploy models to all regions. If you select a single region plan, latency-based routing will not be included.

  • Can I get refunded?

    You may cancel your account at any time, but we do not provide refunds.