
Software Engineer @raycast
Staffordshire, UK
Joined April 2012
- Tweets7.6K
- Following954
- Followers1.2K
- Likes48.6K
Last week I had 2,091 open tabs in Chrome.
430 turned out to be exact duplicates.
My Mac became unresponsive.
@hugorcd taught me his ways.
Every Friday I now clear out my tabs.
Having @raycast on Windows makes it actually usable
AI, Dictation, and Hugo, all in one video?! 🤯
In this What's in my Raycast, @hugorcd shows how he tests MCPs, runs automations every morning, dictates in French and gets English, and asks Raycast AI to build his own extensions.
A must watch!
youtu.be/GpsbbaQvm-I
If you're not using dictation in @raycast yet, why not?
Whether you're running into issues, or missing features from your soon-to-be-ex dictation app, let us know and we'll sort it out.
manual.raycast.com/ai/dictat…
Faster. Lighter. Smoother.
Another week, another round of improvements:
📁 File Search: up to 6× faster results
🧩 Extensions: installs and updates about 4× faster
🧠 Startup: about 20% less memory
📜 Scrolling: smoother than ever
📈 ...and over 40 fixes
👉 raycast.com/changelog
what a combo
🥖 The New Standard is coming to Paris!
October 29th from 6-10pm.
Discussion panel:
- @thomaspaulmann from Raycast
- @artman from @linear
- Théo Carrive from @MistralAI
Spots are limited.
RSVP here → luma.com/4pn0klwq
Nice way to show how your choice determines which sort of models are used.
for the new Auto model in @raycast I came up with a spider/radar chart to allow users to compare the 3 preferences we offer: lightweight, balanced, and flagship. it's a tradeoff between speed, cost, and intelligence. it felt natural and fun to present the capabilities this way. wdyt?
So many improvements shipped, more still to come!
Second round of performance improvements is out:
🧠 Launcher: up to 35% less idle memory
📁 File Search: up to 80% less memory and CPU
⚡ AI requests: fewer credits, faster local models
Plus a few more goodies:
🎯 Send Element to AI
🐞 30+ fixes
👉 raycast.com/changelog
Love this small detail in @Cloudflare, no need to manually type out strings, just click and paste.
We don't show off bug fixes enough, love to see it!
I fixed something that was really bothering me in @raycast where deleting a hotkey from Root Search looked flickery, quickly flashing a few states in a row whilst the action panel closed.
Much more solid now! Will be out in this week's release.
Before/After screen recordings:
Me with my 1 Apple Shortcut 😂
A good case study in how, no matter how much you test, some users will have setups you never covered.
The only way to get everything about Apple Shortcuts (icons, whether they take input) is Apple Events, and fetching them one shortcut at a time cost ~60ms per shortcut.
@stephenrobles has 2400 shortcuts. At 60ms each that's 144s, past our 120s timeout, so he saw nothing.
Batching plus cheaper icon conversion got us to ~6ms per shortcut, and ~1ms once icons are cached.
The attention to detail 👨🍳
As part of this week's release of @raycast I shipped a bunch of improvements related to colour contrast adjustments.
Let's look at the challenges Raycast faces and some examples of how we tackle them…
Fellow dictation user!
Our first What's in my Raycast in V2 is out! @ryannystrom had a chat with @peduarte to showcase his setup and learn some tips along the way. Check it out on YouTube: youtu.be/k6WRf1S9z4M
Being British means appreciating the little things in life, like a proper cup of tea, a polite queue, and lower memory usage in Raycast.
This week’s update is live:
💾 Lower memory usage
🤖 New Auto model stretches your credits further
📊 Track Claude + ChatGPT usage
⌨️ No more dropped keystrokes after hotkeys
🐞 50 more fixes
More performance and memory gains soon.
Changelog: raycast.com/changelog














