# Modal

38 Hacker News commenters recommended Modal and 12 criticized it, November 2025 to November 2026. It was recommended in 11 of 13 months. It meets the Hacker Approved badge rule.

**Verdict:** Recommended for easy serverless GPU and Python compute, sandboxes and endpoints, billed by execution time. Critics mention slow cold starts and outages.

## Why people recommend it

- Strong GPU serverless SDK and API (3 comments): "I think they got FaaS for GPU exactly right, both in terms of their SDK and the abstractions/infra they provide." ([source](https://news.ycombinator.com/item?id=47085541)) "seems like a really nice emerging pythonic way to call out to remote infrastructure" ([source](https://news.ycombinator.com/item?id=46873718)) "Modal has a well documented API, it should be quite straight-forward" ([source](https://news.ycombinator.com/item?id=46081048))
- Pay per execution, predictable costs (3 comments): "You pay for execution time rather than tokens so the unit economics are deterministic" ([source](https://news.ycombinator.com/item?id=46785143)) "renting a cloud instance at Modal and running a model off hugging face is relatively affordable" ([source](https://news.ycombinator.com/item?id=46234606)) "it worked perfect at price point I found reasonable" ([source](https://news.ycombinator.com/item?id=46724037))
- Used by well-known AI companies (2 comments): "Modal builds the serverless cloud used by many of the most popular AI companies you've heard of." ([source](https://news.ycombinator.com/item?id=46872403)) "builds the serverless cloud used by many of the most popular AI companies you've heard of" ([source](https://news.ycombinator.com/item?id=48363551))
- Good for ephemeral sandboxes (2 comments): "discovering how Ramp team use Modal for sandboxed dev environment just saved us weeks of custom infra development" ([source](https://news.ycombinator.com/item?id=46617644)) "Modal for ephemeral sandboxes" ([source](https://news.ycombinator.com/item?id=45798968))

## What people warn about

- Slow cold starts (3 comments): "Cold boot times are around 5m but if your usage periods are predictable it can work out ok." ([source](https://news.ycombinator.com/item?id=47192388)) "cold start was what eventually pushed me off modal for a request-response" ([source](https://news.ycombinator.com/item?id=47758193)) "not having the horrible cold starts of Modal" ([source](https://news.ycombinator.com/item?id=47357256))
- Outages and reliability problems (2 comments): "This is the third major outage in a month." ([source](https://news.ycombinator.com/item?id=48384188)) "Outage on Modal because of this. Not good." ([source](https://news.ycombinator.com/item?id=48057403))
- Locked into platform, limited customization (2 comments): "Cloud sandboxes (E2B, Modal, etc.) work, but you're locked into their platform with limited customization" ([source](https://news.ycombinator.com/item?id=46251791)) "I wrote this originally to replace Modal and Baseten, which I wasn't too pleased with" ([source](https://news.ycombinator.com/item?id=49620842))
- Performance behind rival platforms (2 comments): "we seem to perform better than Modal by ~20% in 4/6 workloads we tested" ([source](https://news.ycombinator.com/item?id=48749758)) "anything compute related just feels like it would have some issues performance wise" ([source](https://news.ycombinator.com/item?id=46570285))

## Alternatives hackers recommend

- [Vast.ai](https://hackerapproved.com/p/vast-ai.md): 37 recommend
- [RunPod](https://hackerapproved.com/p/runpod.md): 28 recommend
- [Verda GPU cloud](https://hackerapproved.com/p/verda.md): 10 recommend
- [Lambda GPU Cloud](https://hackerapproved.com/p/lambda.md): 9 recommend
- [Nebius GPU rental](https://hackerapproved.com/p/nebius.md): 5 recommend

## Where to get it

- [Visit the official site](https://modal.com/)

## Cite as

Hacker Approved, "Modal", https://hackerapproved.com/p/modal/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).