# Max vs wasm inference runtimes

**URL:** https://forum.modular.com/t/max-vs-wasm-inference-runtimes/1459
**Category:** MAX
**Created:** [May 15, 2025, 8:44pm UTC](https://forum.modular.com/t/max-vs-wasm-inference-runtimes/1459 "2025-05-15T20:44:56Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![faramir](https://avatars.discourse-cdn.com/v4/letter/f/e19adc/32.png) [@faramir](https://forum.modular.com/u/faramir)
#### Post date: [May 15, 2025, 8:44pm UTC](https://forum.modular.com/t/max-vs-wasm-inference-runtimes/1459/1 "2025-05-15T20:44:56Z")

</div>

Was watching the latest wasm io conference video about AI and this portion got me curious:

[![](https://us1.discourse-cdn.com/flex001/uploads/modular/original/1X/9784c4e02b2b4b9be7a3d993e2bf1765fad9d7c1.jpeg "Running AI workloads with WASM - is it production ready? [Panel] @ Wasm I/O 2025") ](https://www.youtube.com/watch?v=UODo4Q5ykd8&t=1087)

I didn’t realize that something like wasm could possibly compete with Max. I’m assuming max would be faster but by how much? Wanted to see if anyone else has been thinking about the tradeoffs as there are a lot of pros for wasm. Besides ai, wasm is the next upcoming tech I am very excited for.

---

<div class="post-metadata">

### Author: ![system](https://us1.discourse-cdn.com/flex001/uploads/modular/original/1X/2751e0fbdc595a99718b216730957e9db4448cfd.jpeg) [@system](https://forum.modular.com/u/system)
#### Post date: [November 11, 2025, 8:45pm UTC](https://forum.modular.com/t/max-vs-wasm-inference-runtimes/1459/2 "2025-11-11T20:45:15Z")

</div>

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.