Skip to content

Latest commit

 

History

History
149 lines (69 loc) · 3.22 KB

SITE.md

File metadata and controls

149 lines (69 loc) · 3.22 KB
Raw

Model Deployer

Deploy, manage and rate-limit AI models in production

[Deploy a Model]

[DEPLOY AI MODEL CAROUSEL]

// callout

[MODEL DEPLOYER]

[Deploy On Your Server]

or

[Signup for Model Deployer]

  • A single API for dozens LLMs and Embeddings
  • Supports OpenAI, Anthropic, Mistral, Llamafile and more
  • Track API usage for individual users
  • Charge users by usage or subscription
  • Rate-limit users
  • Offer free faucet for new users
  • Allow your users to provide their own API key

Model Deployer makes it easy to use AI in your apps

Any app developer that wants to use AI models in their apps has a few questions to answer:

Do you charge a subscription? Do you charge based on usage? Do your users signup with their API key? Do you have a free tier? How do you rate limit users and prevent someone from running up your bill? What if your users want to use their own models for privacy?

Model Deployer was created to solve all of these problems and more.

AI models are great, but using them in your app can be tricky because they aren't free! If you run a high-magin SaaS business, you can probably fold it into your costs.

But all kinds of user-facing apps struggle with the economics of using expensive AI models.

Model Deployer solves these problems, by making it easy to use lots of different models in the same interface, tracking their usage, rate limiting, and offering ways to charge your users.

But we've gone even further, because Model Deployer is 100% MIT open-source software, and you can run it yourself for your users. Your users can even run—giving everyone in the ecosystem options.

// feature callout

Manage AI models

Track history, monitor usage and rate limit users in production

Flexible Billing

Give your users the option: monthly subscription, usage based, provide their own API key, or run their own model

Best Large Language Models

All of the major LLMs are supported in a single interface

Text Similarity Embeddings

Multiple Embedding Models are supported in a single interface.

Don't get locked in

Switch between providers with a one-line config change—no code changes necessary.

Open Source

Model Deployer is 100% open-source under the MIT license

// callout

Model Deployer gives you and your users choices:

  1. Use the Model Deployer hosted service to instantly use dozens of AI models
  2. Host Model Deployer yourself and manage which AI models you support
  3. Let your users run their own Model Deployer locally, for maximum privacy

Model Deployer was built on the idea that AI models shouldn’t lock us in. It let’s you trade convenience for privacy—and choose whichever option you and your users prefer.

Deploying Models in Production

Model Deployer makes it easy to deploy dozens of popular AI models in production

Creating models

[SCREENSHOT ADMIN]

Tracking usage

[SCREENSHOT ADMIN]

Viewing history

[SCREENSHOT ADMIN]

Start

Self Host

  • Run Model Deployer on your server
  • Configure which models are enabled
  • Add new local models with Llamafile
  • Track usage
  • Rate limit users

Model Deployer

  • Instantly access dozens of AI models (LLMs, Embeddings)
  • Switch providers with a click
  • Track usage
  • Rate limit users
  • Charge users based on flexible billing model