I appreciate the Fly.io team’s enthusiasm and am optimistic this will mature into a product I’d pay for, but my initial impression was of a lack of polish.
Documentation is sparse, or not even available? The API docs don’t tell you much about the service itself, and a Google search for docs returns an inaccessible website as the first result (https://docs.sprites.dev). Blog posts and forum threads and Claude skills shouldn’t be a substitute.
The snappiness of the sprites is very cool and I can definitely see it integrating into future Claude Code workflows. But the lack of a base container images means you’re still doing setup work on the sprite before you can begin development. I understand the philosophy is that sandboxes should be persistent, but Claude Code sessions also work better when isolated from each other, so it’d be nice to have some precepts to get a workspace setup quickly (given agentic coding is clearly a target).
I also found the CLI unintuitive but maybe that was just me!
So very cool idea but left with the impression that the Fly.io team’s should have spent a couple weeks on polish before shipping.
You're not wrong. The documentation actually had a hallucinated link to an Anthropic dependency in it when we shipped. Right now the attitude is mostly "if we have to document it extensively, we're doing something wrong". It's been in the works for awhile, with a small team, and we're just getting it out there right now.
I've been needling Kurt for several months now that if we wait until it's polished enough that we don't see comments like this, we're doing it wrong.
I'm sure this is a difference-of-learning or whatever, but I'm usually unwilling to try a product until I can understand it and how it works from the documentation
Understandable. Our current take is that there's not really much to know, and that the people this will really light up are good with that. Of course, we'll flesh out documentation!
I'm really jazzed about this particular product as a product (I just really enjoy using it), but the post is mostly about how we built it, and deliberately not much about how best to use it.
I hate being negative but it sounds like par for the course for fly.
Incredible (truly, incredible, world-class) engineers that somehow lack that final 10% of polish/naming/documentation that makes things...well, seriously usable.
I remember last time I tried them the bizarre hoops/documentation around database creation. I _think_ they solved that but I remember at the time it felt almost like I was getting looked down upon as a user. Ugh, you need clarity? how amateurish!
Sounds neat! Also kind of funny, because I already use Fly machines as a bit of a playground for some personal projects and really like it for that purpose. I have minor issues, some of which this blog post directly talks about, but nothing that has held me back.
Also I like a lot of what you get built-in with Fly machines(ex:Grafana). So I am curious how this would fit in my workflow. Maybe Sprites for rapid development and toy projects, and move to a Fly machine when I need something a little more beefy?
I created a "mcp server" for sprites yesterday through this new ecosystem I am working on, you can clone the collection, and then just add this https://tpmjs.com/api/mcp/ajax/sprites/sse mcp sse url to claude desktop, or anything you want.
tpmjs is a registry of ai sdk npm packages (i created them for sprites), which you can add to personal collections, we automatically server your collections as mcp servers if you want.
Tried it. Docker wasn't preinstalled so asked Claude to do it and make sure it's running (supervisord or the likes isn't preinstalled).
It neatly did so and "registered" it as a sprite service. Then I exited my session, waiting for the sprite to go idle, but I don't think it ever does.. Still have it active. Don't know how to idle it.
Can't tell for sure if this means I'm losing credits as there is no billing usage shown anywhere.
Also waiting for the moment where I can launch a sprite from another's checkpoint.
They're hair-trigger inactive otherwise. They don't bill CPU unless they're active. The idea is that there isn't really any uncertainty about when it's running; when you stop interacting with it it stops metering.
This is a new shape for a cloud computing thingy and there'll be snags this week with it, but we don't make our money by billing people for stuff they don't want. We've always gone out of our way not to nickel-and-dime casual users and we're trying hard to find new ways to lean into that here.
(Destroying a Sprite you're done with is a perfectly reasonable move; they're disposable.)
I think the idling feature still needs some work. I created one over the weekend that hasn't idled once, and I've run several tests with sprites that have nothing in them—just `sprite create` and log out, just to see what happens (which unfortunately is nothing, left alone it keeps on running as well.)
I love the idea and most of the execution, I've really enjoyed getting my first sprite configured just the way I want it. It just needs the idling feature to work as advertised before I think I can use it as cost-effectively as it promises.
We had a bug where some sprites would fail to properly suspend while entering their suspended state. You're not eating into credits so no worries there. We've been rolling out a fix across the fleet today so you should be seeing proper status soon.
This is great news! If we upgraded our sprite already how long should it take to suspend? I noticed the upgrade earlier and installed it but my sprite is still running.
ahh finally success—a fresh sprite goes to sleep as it should. unfortunately the original one i created doesn't, so I guess I'm going to have to kill that one off.
I've been having fun with OpenCode's webui in a Sprite, set it up right in Termux and you can vibe code a website with a full backend from your phone. I have a Termux shortcut that sets up a port forward and pops open my web browser to the OpenCode webui. OpenCode can pick up the Sprite skill built into the containers so it can manage the instance itself (mostly setting up and destroying services).
Of course you can do similar with a VPS, but this makes it very easy (and cheap!) to spin up an entire new machine with a public url and HTTPS whenever you want to.
Regardless of the usage of the term Sprite, the real measure of how appropriate it is to use the term for something else is how many people get confused in this manner. I can't really tell what the average reader would think because my background is in game development, so my view is not representative.
I think people can get bogged down in the technical weeds over what a sprite is in graphics. Historically it started out as mini graphics overlays in hardware. There was a transition period motivated by Amiga documentation to have Sprites and Bobs, to distinguish, and perhaps advertise, the use of the Blitter. When software or Blitter Sprites became nearly ubiquitous, they returned to simply being called Sprites, the fairly rare use of the original form became known as Hardware Sprites. Usually it was only mouse pointers that remained as Hardware Sprites
Obviously the term Hardware Sprites is not strictly a distinguishing label either. They are all controlled by software using hardware with some degree of balance between the two.
It’s not even really outdated or unused. Just about everyone who makes 2D games still uses the term “sprite” all the time. The meaning has become slightly less specific, but otherwise is really the same.
I wish I'd had more space to write about the global orchestrator design, because it's fun.
The Fly Machines orchestrator goes through some trouble to keep the source of truth for each VM decentralized, owned by the physical it runs on. But there's still global state --- apps, organizations, services. That stuff is all on Postgres. Postgres keeps up with it just fine but I'd be lying if I didn't say we're always looking out the corner of our eyes on metrics.
The global state for Sprites is on object storage. Each organization gets a separate SQLite database, and that database is synchronized to object storage with Litestream.io (Lightstream is load bearing in a bunch of places here; solid as a rock for us).
I think people really still sleep on the "multiple SQLite database" backing store design.
I've been working on the orchestrator side with Elixir and Phoenix, so happy to continue the discussion for curious minds. One of the coolest things we can do is things like this in Elixir - from any node we can reach out to a sqlite db across the planet:
OrgTracker.with_repo(org_id, fn ->
repo.all(from sprite in "sprites", select: ...)
end)
That will find or place an Elixir process on the cluster and rpc the target node with our code. Placements can be sticky so they pin to a machine so we don't have to suck down the db every start, but we also balance out the load and handle failover of durable processes automatically. Combined with litestream, the result is distributed sqlite with failover while treating it essentially like a locally reachable sqlite db. Yes there is the speed of light to contend with, but by sending the execution across the wire rather than individual queries, we only ever pay a single hop to reach the process/sqlite.
To me, this sounds more monolithic than containers. I think I'd like something less monolithic. However, those who like monorepos might be more quick to develop an interest in this. I could of course use containers within the MicroVM, which is what I really want anyways, because I want lighter weight containers than MicroVMs for sandboxes.
While looking into giving fly another shot as a cloud provider even though I think it's still pretty much a commodity for me, I found an issue in Google: I searched for "fly.io sao paolo" and the title of the first result on fly.io is "Regiones · Fly Docs", translated from english to Spanish. While I find the translation in titles on Google annoying, I haven't often seen the characters messed up like this. I reproduced this in Incognito at this URL: https://www.google.com/search?hl=es&q=fly.io%20sao%20paolo
It's a hardware-isolated Linux shell with a durable disk you can conjure out of the sky on 1.5 seconds notice that costs virtually nothing when you're not actively using it. If we could have shipped this in 2021, before anyone thought coding agents would work, we'd have barfed on our shoes with excitement.
I guess I can see how pre-installing some LLM agents makes it potentially seem "AI-centric", but I don't understand at all how this could be "US-centric".
Thought this was going to be a deep dive into sprite sheets.
Low and behold, yet another AI product selling advertisement + veiled promotion on HN.
It's almost as if the guidelines don't matter anymore. From [0].
'Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.'
[delayed]
Recent and related: Code and Let Live - https://news.ycombinator.com/item?id=46557825 - Jan 2026 (186 comments)
I appreciate the Fly.io team’s enthusiasm and am optimistic this will mature into a product I’d pay for, but my initial impression was of a lack of polish.
Documentation is sparse, or not even available? The API docs don’t tell you much about the service itself, and a Google search for docs returns an inaccessible website as the first result (https://docs.sprites.dev). Blog posts and forum threads and Claude skills shouldn’t be a substitute.
The snappiness of the sprites is very cool and I can definitely see it integrating into future Claude Code workflows. But the lack of a base container images means you’re still doing setup work on the sprite before you can begin development. I understand the philosophy is that sandboxes should be persistent, but Claude Code sessions also work better when isolated from each other, so it’d be nice to have some precepts to get a workspace setup quickly (given agentic coding is clearly a target).
I also found the CLI unintuitive but maybe that was just me!
So very cool idea but left with the impression that the Fly.io team’s should have spent a couple weeks on polish before shipping.
You're not wrong. The documentation actually had a hallucinated link to an Anthropic dependency in it when we shipped. Right now the attitude is mostly "if we have to document it extensively, we're doing something wrong". It's been in the works for awhile, with a small team, and we're just getting it out there right now.
I've been needling Kurt for several months now that if we wait until it's polished enough that we don't see comments like this, we're doing it wrong.
I'm sure this is a difference-of-learning or whatever, but I'm usually unwilling to try a product until I can understand it and how it works from the documentation
Understandable. Our current take is that there's not really much to know, and that the people this will really light up are good with that. Of course, we'll flesh out documentation!
I'm really jazzed about this particular product as a product (I just really enjoy using it), but the post is mostly about how we built it, and deliberately not much about how best to use it.
I hate being negative but it sounds like par for the course for fly.
Incredible (truly, incredible, world-class) engineers that somehow lack that final 10% of polish/naming/documentation that makes things...well, seriously usable.
I remember last time I tried them the bizarre hoops/documentation around database creation. I _think_ they solved that but I remember at the time it felt almost like I was getting looked down upon as a user. Ugh, you need clarity? how amateurish!
Naming? We got naming wrong?
Could not have illustrated the point better if I tried.
Likewise!
Sounds neat! Also kind of funny, because I already use Fly machines as a bit of a playground for some personal projects and really like it for that purpose. I have minor issues, some of which this blog post directly talks about, but nothing that has held me back.
Also I like a lot of what you get built-in with Fly machines(ex:Grafana). So I am curious how this would fit in my workflow. Maybe Sprites for rapid development and toy projects, and move to a Fly machine when I need something a little more beefy?
Looking forward to checking this out.
I created a "mcp server" for sprites yesterday through this new ecosystem I am working on, you can clone the collection, and then just add this https://tpmjs.com/api/mcp/ajax/sprites/sse mcp sse url to claude desktop, or anything you want.
tpmjs is a registry of ai sdk npm packages (i created them for sprites), which you can add to personal collections, we automatically server your collections as mcp servers if you want.
Creating sprites in an example chat bot -> https://imgur.com/a/ETNxR1o
Creating sprites in claude desktop -> https://imgur.com/myC0U28
Listing out my sprites in claude desktop -> https://imgur.com/rgBU0jm
---
You can view the collection of tools here -> https://tpmjs.com/ajax/collections/sprites (fork it to use it yourself)
I'm looking into exe.dev and sprites.dev to build out extra features into tpmjs, agent sandboxes make a lot of sense.
Tried it. Docker wasn't preinstalled so asked Claude to do it and make sure it's running (supervisord or the likes isn't preinstalled).
It neatly did so and "registered" it as a sprite service. Then I exited my session, waiting for the sprite to go idle, but I don't think it ever does.. Still have it active. Don't know how to idle it.
Can't tell for sure if this means I'm losing credits as there is no billing usage shown anywhere.
Also waiting for the moment where I can launch a sprite from another's checkpoint.
There's no way to stop sprites from the CLI.
Supposedly they auto-stop when inactive.
But I've tried multiple times and they don't stop, and it's not just Docker that prevents them from stopping.
I created a new sprite and installed ffmpeg. Then exited. Next day I run `sprite ls` and it's been running continuously for 23 hours.
No way to tell if I'm being billed for it or not.
And the per-hour pricing is extremely expensive.
So for now it's `sprite destroy -s spritename`.
Maybe I'll check it out again in a few months after the fly team has iterated on this a few times.
Sprites are active when:
* They're servicing an incoming HTTP request.
* You're interacting with a console.
They're hair-trigger inactive otherwise. They don't bill CPU unless they're active. The idea is that there isn't really any uncertainty about when it's running; when you stop interacting with it it stops metering.
This is a new shape for a cloud computing thingy and there'll be snags this week with it, but we don't make our money by billing people for stuff they don't want. We've always gone out of our way not to nickel-and-dime casual users and we're trying hard to find new ways to lean into that here.
(Destroying a Sprite you're done with is a perfectly reasonable move; they're disposable.)
No console activity, no HTTP requests, but it doesn't stop.
Can't find any place in CLI or web UI to see how many minutes are charged for CPU, memory, storage.
I think the idling feature still needs some work. I created one over the weekend that hasn't idled once, and I've run several tests with sprites that have nothing in them—just `sprite create` and log out, just to see what happens (which unfortunately is nothing, left alone it keeps on running as well.)
I love the idea and most of the execution, I've really enjoyed getting my first sprite configured just the way I want it. It just needs the idling feature to work as advertised before I think I can use it as cost-effectively as it promises.
There needs to be a way to see how much it is being used then and not simply the life of the Sprite.
For what it's worth, CPU pricing is based on CPU util. A sprite sitting idle CPU costs almost nothing, even when "active".
Yeah. My sprites never idled inspite of having nothing running in them and had to be destroyed. Ideally there should be two settings
1. A timeout after the last console session is exited 2. Force idle using the CLI
Just tried it again. Creation took a very long time, then errored.
But despite the error, the sprite was apparently created, and it runs continuously.We had a bug where some sprites would fail to properly suspend while entering their suspended state. You're not eating into credits so no worries there. We've been rolling out a fix across the fleet today so you should be seeing proper status soon.
This is great news! If we upgraded our sprite already how long should it take to suspend? I noticed the upgrade earlier and installed it but my sprite is still running.
ahh finally success—a fresh sprite goes to sleep as it should. unfortunately the original one i created doesn't, so I guess I'm going to have to kill that one off.
Just use container-os as your runtime image: https://hub.docker.com/r/miget/container-os
and you should be good
Could you explain more about how this helps and how one might use a different runtime image on Sprites?
I don't think you can choose the runtime image on Sprites, not sure what they are referring to here.
I've been having fun with OpenCode's webui in a Sprite, set it up right in Termux and you can vibe code a website with a full backend from your phone. I have a Termux shortcut that sets up a port forward and pops open my web browser to the OpenCode webui. OpenCode can pick up the Sprite skill built into the containers so it can manage the instance itself (mostly setting up and destroying services).
Of course you can do similar with a VPS, but this makes it very easy (and cheap!) to spin up an entire new machine with a public url and HTTPS whenever you want to.
What's the option for iOS? I've tried installing ish and a-shell and can't get the sprites CLI installed on either of them.
Vibe code a client in bash or Python that uses the Sprites HTTP API maybe?
The sprite is a pretty outdated thing these days, but is the term sufficiently unused to make it free for reuse?
I thought this was a post about graphics.
What's a sprite?
https://en.wikipedia.org/wiki/Sprite_(computer_graphics) is the one I'm thinking of.
A graphical object composited just in time in display scanning process.
(Or in other words, a graphical object on screen that is not present in a framebuffer of any kind.)
There was a time before 8bit video games.
https://www.merriam-webster.com/dictionary/sprite
I personally associate the term Shakespeare's The Tempest: https://en.wikipedia.org/wiki/Ariel_(The_Tempest)
This is immediately what popped into my head, and I clicked the link thinking it was going to be some old-school game dev content.
Regardless of the usage of the term Sprite, the real measure of how appropriate it is to use the term for something else is how many people get confused in this manner. I can't really tell what the average reader would think because my background is in game development, so my view is not representative.
I think people can get bogged down in the technical weeds over what a sprite is in graphics. Historically it started out as mini graphics overlays in hardware. There was a transition period motivated by Amiga documentation to have Sprites and Bobs, to distinguish, and perhaps advertise, the use of the Blitter. When software or Blitter Sprites became nearly ubiquitous, they returned to simply being called Sprites, the fairly rare use of the original form became known as Hardware Sprites. Usually it was only mouse pointers that remained as Hardware Sprites
Obviously the term Hardware Sprites is not strictly a distinguishing label either. They are all controlled by software using hardware with some degree of balance between the two.
It’s not even really outdated or unused. Just about everyone who makes 2D games still uses the term “sprite” all the time. The meaning has become slightly less specific, but otherwise is really the same.
I wish I'd had more space to write about the global orchestrator design, because it's fun.
The Fly Machines orchestrator goes through some trouble to keep the source of truth for each VM decentralized, owned by the physical it runs on. But there's still global state --- apps, organizations, services. That stuff is all on Postgres. Postgres keeps up with it just fine but I'd be lying if I didn't say we're always looking out the corner of our eyes on metrics.
The global state for Sprites is on object storage. Each organization gets a separate SQLite database, and that database is synchronized to object storage with Litestream.io (Lightstream is load bearing in a bunch of places here; solid as a rock for us).
I think people really still sleep on the "multiple SQLite database" backing store design.
I've been working on the orchestrator side with Elixir and Phoenix, so happy to continue the discussion for curious minds. One of the coolest things we can do is things like this in Elixir - from any node we can reach out to a sqlite db across the planet:
OrgTracker.with_repo(org_id, fn ->
end)That will find or place an Elixir process on the cluster and rpc the target node with our code. Placements can be sticky so they pin to a machine so we don't have to suck down the db every start, but we also balance out the load and handle failover of durable processes automatically. Combined with litestream, the result is distributed sqlite with failover while treating it essentially like a locally reachable sqlite db. Yes there is the speed of light to contend with, but by sending the execution across the wire rather than individual queries, we only ever pay a single hop to reach the process/sqlite.
For those who missed it, tptacek wrote TFA.
>Any time anybody needs a pen, several are at hand, in exactly the first place they look.
>Now, today, under the hood, Sprites are still Fly Machines. But they all run from a standard container.
So its basically a faster VM?
What are some use cases for this that benefit from the faster boot time?
Is there a way to copy files to/from the sprite? Couldn't see an option for that in the CLI
There's an FS API that shipped yesterday:
It'll get CLI support very soon.But I've just been using magic-wormhole for this.
How does Sprites compare to E2B sandboxes?
To me, this sounds more monolithic than containers. I think I'd like something less monolithic. However, those who like monorepos might be more quick to develop an interest in this. I could of course use containers within the MicroVM, which is what I really want anyways, because I want lighter weight containers than MicroVMs for sandboxes.
While looking into giving fly another shot as a cloud provider even though I think it's still pretty much a commodity for me, I found an issue in Google: I searched for "fly.io sao paolo" and the title of the first result on fly.io is "Regiones · Fly Docs", translated from english to Spanish. While I find the translation in titles on Google annoying, I haven't often seen the characters messed up like this. I reproduced this in Incognito at this URL: https://www.google.com/search?hl=es&q=fly.io%20sao%20paolo
[flagged]
It's a hardware-isolated Linux shell with a durable disk you can conjure out of the sky on 1.5 seconds notice that costs virtually nothing when you're not actively using it. If we could have shipped this in 2021, before anyone thought coding agents would work, we'd have barfed on our shoes with excitement.
I guess I can see how pre-installing some LLM agents makes it potentially seem "AI-centric", but I don't understand at all how this could be "US-centric".
the US is the centre of the world, and AI is at the centre of US attention!
why would anyone do anything else, anywhere else?
This is not the kind of sprites I'm interested in
Thought this was going to be a deep dive into sprite sheets.
Low and behold, yet another AI product selling advertisement + veiled promotion on HN.
It's almost as if the guidelines don't matter anymore. From [0].
'Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.'
In fairness, I don't think the poster is a Fly.io employee, so that guideline doesn't quite apply here.