WASP — Run PHP (WordPress, Laravel, custom apps) natively on the Internet Computer
WASP (WASm-Php) will package a PHP 8-compatible runtime inside an ICP canister so teams can deploy WordPress, Laravel, and custom PHP directly on-chain. It will ship with WASQL (multi-user SQL engine) and a stable-memory filesystem, allowing mainstream PHP apps to run without a rewrite.
Planned status at first public preview: ~70% overall interpreter coverage.
Why this will matter (market context)
PHP powers ~73.6% of websites that use a known server-side language. (w3techs.com)
WordPress runs ~43.4% of all websites (≈60.8% CMS share). (w3techs.com)
WooCommerce accounts for ~6.0M live sites on WordPress. (BuiltWith)
Laravel is detected on ~749k live sites (BuiltWith); Wappalyzer tracks ~125k (different crawler scope). (BuiltWith, wappalyzer.com)
What will ship
1) WASP (PHP-in-a-Canister Runtime)
A PHP interpreter embedded in a canister, mapping PHP I/O, time, and networking to ICP-safe primitives. It will serve certified HTTP and integrate Internet Identity for authentication.
2) WASQL (Multi-User SQL for ICP)
An in-canister SQL layer with SQLite-style semantics (transactions, checkpointing), adapted to ICP’s update/query model for multi-tenant apps (e.g., WP multisite, Laravel SaaS).
3) Stable Filesystem (stable-fs)
An application-level filesystem backed by stable memory (paths, metadata, streaming reads/writes, upgrade-safe persistence) so uploads, plugins, and themes behave predictably.
do you have any ETA for the first public preview? are you solely working on this? if so, do you have any plans to involve a broader dev community in this project?
I kindly also inform you about the Call for Ecosystem Updates thread where you can share updates to be included in the dev newsletter.
FYI → you already got quite some attention on socials
For your questions, no timeline for now, I’m working hard to get all the official PHP tests to pass, but I’d love to have a version of WordPress running by the end of the year This will mainly depend on the 40B operation limit and the strategy I’ve designed (state management).
This is a private project for now; I submitted a grant request that was denied, which would have helped me set up an open-source strategy, a DAO, and an ICO to provide a full professional hosting platform. For now, I just have to keep it simple
For the public preview, there are examples or you can test your own PHP code, just replace the code in the input window :).
So the database would sit inside of a smart contract which is secured by cryptography and is essentially locked up and encrypted and I’m assuming this encryption is superior to simply using SSL on web2 or not? But definitely decentralized, redundant, and divorced from big tech cloud meaning hopefully less or no censorship and no takedowns and “unstoppable”?
So you would want to move your WordPress site to ICP in order to truly secure the data in the database since exploits are much harder due to the smart contract security itself vs. all the doorways with web2 (like phpMyAdmin) but also because you want to utilize web3 and possibly crypto tipping right?
Yeah actually this could be a huge sell and bring lots of sites to ICP and make it extremely easy to port such sites. I’m very surprised you didn’t get a grant for this. This is truly a very good idea. One of the major complaints from companies is that porting over is difficult but this would truly make the transition a lot easier until they decide to do full rewrites from scratch.
What would upgrades be like if needed for the base software here? I’m assuming you’d have to manually push new updates as they become available or could something like a linux package manager pull lets say the latest php update from an external repo into the canister?
Anyways, I hope you bring this to production and resubmit your grant application when the time is right. This is a very good idea if it works.
You are right about all the points mentioned. Migrating sites is not really a problem in itself with the use of plugins, it’s quite easy. However, DNS changes and, let’s say, the fact that email accounts are often tied to the domain complicate the migration.
Regarding updates, only the Core (WASP,the file system and WASQL) and the PHP version are problematic, since everything related to WordPress or Laravel would be handled internally. We will therefore need to include backup mechanisms and treat updates as transactions with the possibility of rollback.
This is very promising, hope you keep working on it.
Out of curiosity when you applied for the grant, did you present some code/PoC? Cause if you did and still got rejected that is quite troubling imho and warrants serious reconsideration of the criteria used to award grants
How is the project going? Saw the last update was in September. How can I best support this project? There have got to be others that are willing to pitch in. Will it be open source?
All the benefits of PHP compatibility and portability secured inside of a smart contract running on the World Computer ICP with the added ability to utilize the internet of value.
DFINITY should grant this project because we want to encourage migration to the IC in the simplest, cheapest, fastest way without forcing massive rewrites from the get-go. It’s a no-brainer strategy to gently transition business over to web3 without forcing them to immediately commit to an entire rewrite and paradigm shift while still reaping some of the benefits.
Correct me if I’m wrong here?
The question then becomes. Which sites would benefit the most?
And, what would the cost be vs. hosting on web2 big tech cloud? With the same bandwidth/traffic?
I’m still working on integrating Laravel and WordPress. The goal is to have Laravel running over the next few weeks, as planned, and WordPress should be easier to finalize afterward. Right now, I’m focused on the autoloader, which is the last missing piece to get Laravel loading correctly.
The most critical aspect at this stage is the number of cycles consumed by each query, as this will determine many architectural choices. For example, if a WordPress request requires more than 40B cycles to execute— which is the maximum for a single update call— I’ll need to implement a state machine, which will take more time to build.
I’ve also run into the maximum memory an ICP Wasm module can use. Switching to wasm64 has given me a bit more headroom. I’m trying to keep everything inside a single canister, since relying on multiple canisters would introduce overhead and slow down execution.
On the SQL side, I’ve made solid progress by adding transactions and multi-user support. WASQL is now more than just a simple SQLite clone.
Regarding open source, my long-term vision has always involved creating a DAO, and I’m still considering that path—for example, a DAO focused on WordPress/Laravel/PHP hosting.
I’ve also started evaluating hosting costs based directly on cycle usage. With the help of ChatGPT, I ran some calculations and assumptions showing how implementing caching can significantly reduce costs.
Excited for this project.
Does this project, once complete, allow to migrate an existing Wordpress project to ICP?
And allow to use any/ most Wordpress plugins?
Tentatively, being able to easily deploy a WordPress site in a canister is the win everyone saying
“Wen Pump?” has been looking for. However, will this architecture be cost prohibitive?
Just added it to 
