Max Platform 

The Modular Accelerated Xecution (MAX) Platform is an integrated suite of tools for AI deployment that power all your AI workloads of today, and tomorrow.

Give your AI developers superpowers

Enable your AI developers to achieve more with less. Empower them to bring the newest AI innovations to production faster, with higher performance and lower cost. All without having to rewrite your stack.

Fewer things that work better

One set of tools to support your entire inference pipeline and all your AI models — reduce complexity and simplify deployment.

Unparalleled performance

Productionize your GenAI and traditional AI inference pipelines within your latency and budget constraints — take full advantage of any hardware out of the box.

Meets you where you are

Drop-in compatible with the tools and technologies you already use in production — realize more value faster without rewriting your stack.

Support all your generative and traditional AI use cases

MAX provides drop-in compatibility with any model from any framework.  Support for all the framework operations, quantized types, dynamics shapes, and your custom operations.

Sign up now

Deploy on the world’s most performant & cost efficient unified infrastructure

MAX unlocks state-of-the-art latency and throughput for your generative and traditional AI models, helping you productionize larger models and realize massive cost savings on your cloud bill.

125 qps
Modular Engine
* Model
AWS c6g.4xlarge
Batch Size

Plug into what you already use

MAX integrates with the your existing AI serving stack so you can deploy to production using the libraries and tools you know and trust. Adopt what's next without throwing away what works.

Learn more

Enterprise-grade AI that meets your security and privacy standards

MAX integrates seamlessly into your enterprise, with superior data protection, privacy, and compliance controls specialized to your industry.

Get started with Mojo 🔥

Download the Modular CLI to download and play with Mojo on your local machine. We currently support Mac, Linux, and Windows (WSL) operating systems.

Unpacking modular (0.1.0) ... Setting up modular (0.1.0) ...
  __  __           _       _
 |  \/  |         | |     | |
 | \  / | ___   __| |_   _| | __ _ _ __
 | |\/| |/ _ \ / _` | | | | |/ _` | '__|
 | |  | | (_) | (_| | |_| | | (_| | |
 |_|  |_|\___/ \__,_|\__,_|_|\__,_|_|

Thanks for installing the Modular CLI!
To get started, type "modular --help".
To install Mojo, type "modular install mojo".

Visit for our getting started guide.
Join our Discord at,
or file issues on Github at
modular install mojo # Found release for @ 0.3.0--release, installing to /root/.modular/pkg/packages.modular.com_mojo
# Downloads complete, setting configs...
# Configs complete, running post-install hooks...
Testing `MODULAR_HOME=/root/.modular` * `/root/.modular/pkg/packages.modular.com_mojo/bin/mojo`...
TEST: `mojo --help`... OK
TEST: `mojo run --help`... OK
TEST: `mojo build test_mandelbrot.mojo`... OK
TEST: `mojo build test_python.mojo`... OK
TEST: `mojo demangle`... OK
reformatted /tmp/tmpmclc50j5/test_format.mojo

All done! ✨ 🍰 ✨
# Installation complete. Mojo is installed under $HOME/.modular. Set $MODULAR_HOME in your current shell to this location, and add $HOME/.modular/pkg/packages.modular.com_mojo/bin to your PATH.
# Downloads complete, setting configs...
# Configs complete, running post-install hooks...
🔥 Mojo installed! 🔥
python3 Throughput of a 128x128 matrix multiplication in Python:
0.00215 GFLOP/s
mojo matmul.🔥 Throughput of a 512x512 vectorized + parallelized matrix multiplication in Mojo:
79.636 GFLOP/s (36870x speedup over Python)

Why Modular?

Built by the world’s AI experts,

Our team has built most of the world’s existing AI infrastructure, including TensorFlow, PyTorch, XLA, and we’ve built and scaled dev tools like Swift, LLVM, & MLIR. Now we’re focused on rebuilding AI infrastructure for the world.

Reinvented from the ground up

To unlock the next wave of AI innovation, we need a “first principles” approach to the lowest layers of the AI stack. We can’t pile on more and more layers of complexity on top of already over-complicated existing solutions.

Built with generality in mind

Natively multi-framework, multi-hardware, and multi-cloud — our infrastructure was designed from the beginning to scale from the largest accelerator clusters down to the smallest edge devices and in-between.

Infrastructure that just works

We build technology that meets you where you are. We don’t require you to rewrite your models, workflows, or application code, grapple with confusing converters, or be a hardware expert to take advantage of bleeding-edge technology.

Built for you

Move beyond Big Tech’s trickle-down infrastructure and the challenges of open-source software. Get direct access to industry experts that can solve any issue you have with our infrastructure and make sure we’re meeting your SLA/SLOs.

Ready to get started?

Sign up to gain access to Modular’s infrastructure.

Read the docs