Enable Any App to Make Intelligent Decisions

Deploy scikit-learn and online machine learning models to 5 data centers around the world with ease. No servers required.

  • Fast, scalable, and available. Latency-routed models with automatic failover. Response times < 60ms, globally.
  • Simple and Serverless. Deploy with a single line of code, let us manage the servers.
  • Create Intelligent Apps. Personalize content, detect malicious users, recommend products, make your smart home devices smarter, and more.

Make your application more intelligent in minutes.

sign up for free or view our pricing plans starting at $10/mo.

Why Us?

Serverless, high-performance machine learning contained in a simple API.

Scikit-Learn Models

Deploy your locally-trained scikit-learn model with a single line of Python. Redeploy models as frequently as needed (e.g., after batch model updates). Receive predictions by calling the mlrequest API. See the docs for examples.

Online Learning Models

Deploy online classification, regression, and reinforcement learning models. Model updates occur incrementally, in real-time. Unlike traditional ML, you will never have to retrieve your entire training dataset to update your models.

Fast, Scalable, and Available

Make thousands of model transactions per second. Models are duplicated across 5 data centers around the world. Less than 60ms response time, globally.
See status.mlrequest.com.

Integrate with Any App

Perfect for websites, web apps, mobile apps, IoT apps, smart home apps, and more. If your app can make an HTTP request, it can be intelligent.

Simple and Serverless

No server deployment, maintenance, or fees. mlrequest packages a highly-scalable, available, and efficient machine learning service into a simple API. Lightweight and dependency free.

Unlimited Model Training

Paid plans receive unlimited online model training transactions. Modify your model as much as you need until it is perfect, without paying extra.

Secure

HTTPS transmission required. Hash your raw data to maximize security with no loss in machine learning performance. Raw data is automatically removed from our servers after a few days.

Support

We are available to help you when you need it. Email support@mlrequest.com for help.

Frequently Asked Questions

Send us an email if you don't find your question here.

  • What is mlrequest?

    mlrequest is a high-performance API for deploying machine learning models that do not require server setup, maintenance, or software dependencies. It's easy to get started, see the docs for examples.

    We are running in data centers around the world to optimize latency and throughput. mlrequest is designed to be fast, scalable, and available so that you can focus on your machine learning-based application.

  • What can I do with mlrequest?

    mlrequest can be applied to many diverse problems across a wide range of industries or research efforts. Integrate with a large production application, personal project, or anything between. A few examples are listed below.

    • Determine user's intent to buy, signup, or click
    • Learn smart home device profiles
    • Detect fraudulent behavior
    • Personalize app content for individual or groups of users
    • Detect bot and other malicious traffic
    • Recommend products
    • Estimate customer lifetime value
    • Optimize email campaigns
    • Determine user sentiment
    • Identify customers who are likely to churn
  • Where can I learn more about using the API? 

    The docs contain details with examples about how to use the API. 

  • I'm new to machine learning, can I still use mlrequest? 

    Yes, mlrequest is well suited for beginners and experts alike. The API and documentation are designed to make machine learning as accessible as possible. We want you to hit the ground running with integrating machine learning into your application. See the docs for examples on how to get started.

  • What's included in the free tier?

    The free tier provides 5,000 prediction (or online training) transactions per month on a single model. Deployed scikit-learn models are limited to 1 MB in size. Response times for free plans are slightly slower than paid plans by as much as 50ms, to prioritize model transactions for paid accounts. The free plan is intended for app development and hobbyist projects.

    If you require more monthly predictions, more models, larger scikit-learn model files, unlimited online model training, or faster response times then please upgrade to one of our paid plans.

  • Where can I request features, report issues, or submit other feedback?

    Please submit any and all feedback using our feedback form.

    We take feedback seriously and encourage you to let us know how we can improve mlrequest.

Get started in a few minutes

 

Completely free for up to 5,000 monthly transactions

Sign up for free

or view our pricing plans starting at $10/mo.