Modal
Hacker ApprovedThe most recommended of 6 GPU clouds
Recommended for easy serverless GPU and Python compute, sandboxes and endpoints, billed by execution time. Critics mention slow cold starts and outages.
Recommended for easy serverless GPU and Python compute, sandboxes and endpoints, billed by execution time. Critics mention slow cold starts and outages.
50 people gave an opinion
- recommend it
- 38
- criticize it
- 12
- approval share
- 76%
Recommended in 11 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: not recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it carries the badge
- 38 people recommend itNeeds 10 or more, met.
- 3.2 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 11 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Strong GPU serverless SDK and API
“I think they got FaaS for GPU exactly right, both in terms of their SDK and the abstractions/infra they provide.”
Comment from 20 Feb 2026 in “Almost Every infrastructure decision I endorse or regret after 4 years” “seems like a really nice emerging pythonic way to call out to remote infrastructure”
Comment from 3 Feb 2026 in “Show HN: Sandboxing untrusted code using WebAssembly” “Modal has a well documented API, it should be quite straight-forward”
Comment from 28 Nov 2025 in “Migrating to Positron, a next-generation data science IDE for Python and R” Pay per execution, predictable costs
“You pay for execution time rather than tokens so the unit economics are deterministic”
Comment from 27 Jan 2026 in “Ask HN: How do you budget for token based AI APIs?” “renting a cloud instance at Modal and running a model off hugging face is relatively affordable”
Comment from 11 Dec 2025 in “The Walt Disney Company and OpenAI Partner on Sora” “it worked perfect at price point I found reasonable”
Comment from 22 Jan 2026 in “Keeping 20k GPUs Healthy” Used by well-known AI companies
“Modal builds the serverless cloud used by many of the most popular AI companies you've heard of.”
Comment from 3 Feb 2026 in “Ask HN: Who is hiring? (February 2026)” “builds the serverless cloud used by many of the most popular AI companies you've heard of”
Comment from 1 Jun 2026 in “Ask HN: Who is hiring? (June 2026)” Good for ephemeral sandboxes
“discovering how Ramp team use Modal for sandboxed dev environment just saved us weeks of custom infra development”
Comment from 14 Jan 2026 in “Why We Built Our Own Background Agent” “Modal for ephemeral sandboxes”
Comment from 3 Nov 2025 in “A Durable Coding Agent – With Modal and Restate”
What people warn about
Each mark is one comment
Slow cold starts
“Cold boot times are around 5m but if your usage periods are predictable it can work out ok.”
Comment from 28 Feb 2026 in “How do I cancel my ChatGPT subscription?” “cold start was what eventually pushed me off modal for a request-response”
Comment from 13 Apr 2026 in “Show HN: Chipmunkify – I used ML to solve audio's dumbest problem” “not having the horrible cold starts of Modal”
Comment from 12 Mar 2026 in “IonRouter (YC W26) – High throughput, low cost inference” Outages and reliability problems
“This is the third major outage in a month.”
Comment from 3 Jun 2026 in “Modal Major Outage” “Outage on Modal because of this. Not good.”
Comment from 8 May 2026 in “AWS EC2 outage in use1-az4 (us-east-1)” Locked into platform, limited customization
“Cloud sandboxes (E2B, Modal, etc.) work, but you're locked into their platform with limited customization”
Comment from 13 Dec 2025 in “BoxLite Love AI agent – SQLite for VMs: embeddable AI agent sandboxing” “I wrote this originally to replace Modal and Baseten, which I wasn't too pleased with”
Comment from 9 Sep 2026 in “Show HN: Self-host open-source LLMs on AWS with scale-to-zero” Performance behind rival platforms
“we seem to perform better than Modal by ~20% in 4/6 workloads we tested”
Comment from 1 Jul 2026 in “Reduce GVisor Cold Starts with GPU Snapshotting” “anything compute related just feels like it would have some issues performance wise”
Comment from 10 Jan 2026 in “GPU Memory Snapshots: Supercharging Sub-Second Startup”
Other GPU clouds people recommend
The four most recommended other GPU clouds, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| ModalThis page | 38 | 12 | 76% | 11 | — | Approved |
| Vast.ai | 37 | 1 | 97% | 9 | — | Approved |
| RunPod | 28 | 5 | 85% | 11 | — | Approved |
| Verda GPU cloud | 10 | 1 | 91% | 4 | — | Approved |
| Lambda GPU Cloud | 9 | 2 | 82% | 5 | — | — |
What people said
40 of the 59 comments that mention it, newest first. Each links to the original comment.
“I wrote this originally to replace Modal and Baseten, which I wasn't too pleased with”
“Modal significantly improves this. Highly recommend.”
“decided to build it myself on modal. it's self hosted so you can tinker with it”
“The agents found a Modal-hosted insecure app with a weak API key”
“they use gVisor and not Firecracker, which means it's harder for users to run docker inside the box.”
“My main issue with Modal is that their autoscaling is not as good as Daytona's.”
“Had I not thought it was cool I would not have written about it”
“its accessible right now on Modal via Endpoints”
“we seem to perform better than Modal by ~20% in 4/6 workloads we tested”
“Major Sandbox providers (e.g. Modal) run on non-hyperscaler bare metal not AWS and so don't need to markup on AWS's markup.”
“it seems Modal will add $70 to your free "standard" accounts when you try out Endpoints”
“This is the third major outage in a month.”
“builds the serverless cloud used by many of the most popular AI companies you've heard of”
“modal.com was the fastest, it still took around a minute-ish”
“about a second per gigabyte of weights, plus a second to start the container”
“Outage on Modal because of this. Not good.”
“cold start was what eventually pushed me off modal for a request-response”
“If you need fast serverless GPU deployments, look into platforms like Modal.”
“not having the horrible cold starts of Modal”
“think they run a great service”
“Cold boot times are around 5m but if your usage periods are predictable it can work out ok.”
“Our one iPhone member is content with Modal”
“I think they got FaaS for GPU exactly right, both in terms of their SDK and the abstractions/infra they provide.”
“Super impressed with what you've all done at Modal!”
“seems like a really nice emerging pythonic way to call out to remote infrastructure”
“Modal builds the serverless cloud used by many of the most popular AI companies you've heard of.”
“have also noticed this pattern in Modal [1] and Ray [2] and all for it.”
“You pay for execution time rather than tokens so the unit economics are deterministic”
“it worked perfect at price point I found reasonable”
“I recommend using modal for renting the metal.”
“discovering how Ramp team use Modal for sandboxed dev environment just saved us weeks of custom infra development”
“making Modal cold boots fast”
“anything compute related just feels like it would have some issues performance wise”
“Modal is great for infrequent batches.”
“it's hard to go wrong using any of the cloud GPU hyperscalers”
“Companies like Modal have great info about how to build agents with sub-second voice-to-voice latency.”
“Cloud sandboxes (E2B, Modal, etc.) work, but you're locked into their platform with limited customization”
“renting a cloud instance at Modal and running a model off hugging face is relatively affordable”
“Modal has a well documented API, it should be quite straight-forward”
“Modal for ephemeral sandboxes”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.