<3 This has been a work of passion for the past two years of my life (off and on). I hope anyone who uses this can feel the love and care I put into this, and subsequently the amazing private beta community (all ~5,000 strong!) that helped improve and polish this into a better release than I ever could alone.
Ghostty got a lot of hype (I cover this in my reflection below), but I want to make sure I call out that there is a good group of EXCELLENT terminals out there, and I'm not claiming Ghostty is strictly better than any of them. Ghostty has different design goals and tradeoffs and if it's right for you great, but if not, you have so many good choices.
Shout out to Kitty, WezTerm, Foot in particular. iTerm2 gets some hate for being relatively slow but nothing comes close to touching it in terms of feature count. Rio is a super cool newer terminal, too. The world of terminals is great.
I’ve posted a personal reflection here, which has a bit more history on why I started this, what’s next, and some of the takeaways from the past two years. https://mitchellh.com/writing/ghostty-1-0-reflection
Just make sure not to get caught in the pitfall that is maximum render speed, which can lead to missing out on efficiency during slow and partial rendering.
Missing damage tracking, always painting everything (even when the window is a full 4k monitor), etc. kills performance and input latency when dealing with realistic redraw workloads like text editing, blinking cursors and progress bars. Much too often to terminals worry only about the performance of `cat /dev/urandom`...
I gave up on alacritty because it was always using the dedicated graphics card of my MacBook and there was no reasonably way to use the integrated graphics card because it was “low performance”.
- Ghostty does vsync by default and supports variable refresh rates (DisplayLink). If you're on battery and macOS wants to slow Ghostty down, it can and we respect it.
- Ghostty picks your integrated GPU over dedicated/external
- Ghostty sets non-focused rendering threads as QoS background to go onto E-cores
- Ghostty slows down rendering significantly if the window is obscured completely (not visible)
No idea if Alacritty does this, I'm not commenting about that. They might! I'm just talking from the Ghostty side.
I, a person who don't care about battery performance one iota (because my computer has no battery), love this answer and approach. Not every software is for everyone, and authors drawing a line in the sand like that works out better for everyone in the long-term, instead of software that kind of works OK for everything.
In some cases yes. In this case, in my opinion it can be strictly wrong.
The GPU requirements of a terminal are _minuscule_ even under heavy load. We're not building AAA games here, we're building a thing that draws a text grid. There is no integrated GPU on the planet that wouldn't be able to keep a terminal going at an associated monitor's refresh rate.
From a technical standpoint, there is zero downside whatsoever to always using the integrated GPU (the stance Ghostty takes) and plenty of upside.
Because _my_ computer has no battery. There is a plethora of computers out there with batteries who can run Linux, Windows, and macOS. These computers can, on paper, run Alacritty.
Cherry on top is me being a former user of a MBP 2010 who'd crash when using discrete GPU (it was _the_ reason Apple went with AMD later on). And some apps insisted on using it, even when I disabled it.
I like Rust applications but I don't like this response. The dev sounds worn out; whereas the dev of Ghostty seems to be a pleasure to deal with.
> - Ghostty slows down rendering significantly if the window is obscured completely (not visible)
About this. For whatever reason, I often end up with foreground windows (e.g. Chrome) covering the background window entirely, except for a sliver a few pixels wide.
Would Ghostty handle this case? I don't believe there's any point in full-speed rendering if less than a single line of text is shown, but the window isn't technically obscured completely.
> - Ghostty sets non-focused rendering threads as QoS background to go onto E-cores
Assuming you're referring to Apple Silicon chips, how does Ghostty explicitly pin a thread to an E-core? IIRC there isn't an explicit way to do it, but I may be misremembering.
But maybe to add a little bit of context, "damage tracking" means for example, that if there is any ongoing animation (like a spinner), then only a small part of the screen will be re-rendered (with proper vertex-time scissors, so only relevant pixels will be computed). I am not sure if it makes sense in the context of a terminal emulator, but it's certainly a big issue for any non-toy GUI toolkits.
GPUs are incredibly fast parallel computers, so you will likely not observe any perf difference (unless you need order-dependent transparency which you don't), but it might improve your battery life significantly.
No, damage tracking is important as it about reporting that you only updated that spinner, which means that your display server also knows to only redraw and repaint that area, and in turn that your GPU knows during scanout that only that area changed.
Without, even if you only redrew your spinner, your display server ends up having to redraw what might be a full screen 4k window, as well as every intersecting window above and below until an opaque surfaces are hit to stop blending.
Well it sounds like ghostty is like all the other major GPU terminal emulators (unless you know of a counterexample) and does full redraw, though it appears to have some optimizations about how often that occurs.
The power issue might be true in some cases, but even as foot’s own benchmarks against alacritty demonstrate, it’s hyperbolic to say it “kills performance”.
We do a full redraw but do damage tracking ("dirty tracking" for us) on the cell state so we only rebuild the GPU state that changed. The CPU time to rebuild a frame is way more expensive than the mostly non-existent GPU time to render a frame since our render pipeline is so cheap.
As I said in another thread, this ain't a AAA game. Shading a text grid is basically free on the GPU lol.
It's actually not at all free, even if you're barely using the shading capacity. The issue is that you keep the power-hungry shader units on, whereas when truly idle their power is cut. Battery life is all about letting hardware turn off entirely.
Also, if you do damage tracking, make sure to report it to the display server so they can avoid doing more expensive work blending several sources together, and in case of certain GPUs and certain scanout modes, also more efficient scanout. Depending on your choice of APIs, that would be something like eglSwapBuffersWithDamage, wl_surface_damage_buffer, and so forth.
> The issue is that you keep the power-hungry shader units on, whereas when truly idle their power is cut.
Even in the most perverse scenario of a single cell update the load for a terminal is still bursty enough that it's not like the GPU doesn't enter some power saving states. Running intel_gpu_top in kytty with a 100ms update is at least suggestive, it never drops below 90% RC6 (even at 50ms, which is a completely uselessly fast update rate we're still in the high 80s). If you're updating faster than 100ms legitimately, it's probably video or animation that is updating a large percentage of the display area anyway. The overall time my terminal is doing some animation while on battery is low enough that in practice it just doesn't matter.
The problem you're up against is that, maybe if this was optimized most people would get 2 or 3 (or even 10) more minutes on a 12 hour battery life or something. No one really cares. Maybe they should, but they don't. And there's plenty of other suck in their power budget.
You make it sound like a binary power saving scenario, but it tends to be more nuanced in practice.
Most people run their terminal opaque, most display systems optimize this common case.
> Also, if you do damage tracking, make sure to report it to the display server
I'm not unsympathetic to your point of view. But I am skeptical that the power savings for most use cases ends up being much of a big deal in practice for most people (even accepting there may be some annoying edge cases that annoy some). I am interested in this topic, but I am still awaiting an example of a GPU accelerated terminal emulator that works this way to even make a real world comparison.
It is very nuanced, but it's important to realize how small the power budget is and how just tens of milliwatt here and there make a huge difference.
To get over 12 hours of battery life out of a 60 Wh battery - which isn't impressive nowadays with laptops rocking 20+ hours - you need to stay below 5 watts of battery draw on average, and considering that the machine will likely do some actual computation occasionally, you'll need to idle closer to 2-3 watts at the battery including the monitor and any conversion losses.
The really big gains in battery life are from cutting tens to hundreds of mW off things at the bottom by keeping hardware off and using fixed function hardware like e.g. avoiding rendering and doing direct scanout and using partial panel self refresh. Execution units do not turn on and off instantly so pinging them even briefly is bad, and the entire system needs to be aligned with the goal of only using them when necessary for them to stay off.
Efforts like libliftoff to do efficient plane offload to avoid render steps in the display server can save in the area of half a watt or more, but it's not a whole lot of help if applications don't do their part.
Bigger GPUs than your iGPU (or even just later iGPUs) will also likely see even bigger impacts, as their bigger shader units are likely much hungrier.
(As an aside, I am not a fan of kitty - they have really weird frame management and terrible recommendations on their wiki. Foot, alacritty or if ghostty turns out good, maybe even that would be better suggestions. Note that comparing to foot can give a wrong image, as CPU-based rendering pushes work to the display server and gives the illusion of being faster and more efficient than it really is.)
Well I would be very interested in some of this, but it all seems theoretical and mythical. Seriously, what terminal is giving 20% better battery life (or whatever number that people will notice) than kitty.
How can I observe any of these claims in practice?
You’ve put some down some bold claims about how things should be done but no way to verify or validate them at all.
Put up with some real power benchmarks or this is just crack pot.
> To get over 12 hours of battery life out of a 60 Wh battery - which isn't impressive nowadays with laptops rocking 20+ hours
I used 12 hours to be nice. The sell of getting another 10 minutes or so out of 20 hours is even more stark.
The cases where you push a line and scroll, you're repainting most of it anyway. The cases where you're not are end up being infrequent enough that optimizing them in the ways suggested makes an unnoticeable impact.
Build it and they will come maybe?
> Bigger GPUs than your iGPU (or even just later iGPUs) will also likely see even bigger impacts.
In most cases people can get by with an iGPU for a battery laptop cases. If you're in a must pull down more graphical power case, you're often plugged in and few care about 10s of milliwatts then.
> (As an aside, I am not a fan of kitty - they have really weird frame management and terrible recommendations on their wiki. Foot, alacritty or if ghostty turns out good, maybe even that would be better suggestions. Note that comparing to foot can give a wrong image, as CPU-based rendering pushes work to the display server and gives the illusion of being faster and more efficient than it really is.)
Once again, what is the exemplar of an efficient terminal then.
We've already established ghostty doesn't operate the way you think it should so how can it turn out good?
Perhaps that came across wrong, no shade intended, what was meant was that ghostty seems to be like all the other mature GPU based emulators. Which means there's no damage reporting to a display server or anything like that. I don't think it's quite the deal breaker the GGP implies.
He mentions input latency[1] as one of 4 aspects of being fast that were considered during development. I’m not aware of how that was tested, but would trust that it outperforms Iterm2 in that regard.
Congrats on your 1.0 release! Building a term from scratch is no small task. And it looks like after 2 years you already have something I'm seriously considering switching to. That's seriously impressive. And though I'm mostly a Linux supremacist, I genuinely appreciate and respect your dedication to providing native applications on all major platforms. I see too many projects do one of two things: only support one platform at first, with some vague promise of doing others in the future left unfulfilled for years. Or going the easy route of using electron or some other cross-platform toolkit, getting cross-platform support at the cost of a significantly worse experience. There is no substitute for doing the extra work, at the end of the day, and you're already well on your way there at such an early point. Again, I'm impressed!
Now, the one thing I've coveted for years is a linux term with tmux control mode support, ever since I used iterm2 during my short tenure as a Mac user. I can see from your github issues that this is in the works(https://github.com/ghostty-org/ghostty/issues/1935), which is great. It looks like a lot of the necessary UI elements are already in place, as well as partial support for the control mode protocol. Do you have any idea how far off this is from being completed? I don't wanna make demands or anything, I know this is a passion project with lots of things still to work on. As far as I know(and I look around quite often) control mode support is still something no existing term for linux supports. Being the first could drive a lot of adoption! I'd also like to offer my help in pushing it over the goal line. I haven't touched Zig, but I'm a polyglot with plenty C experience and Im highly motivated to bring this feature to my Linux systems. Some guidance on things I might do to help would be welcome. The final 3 points of the checklist don't have much detail provided.
Again, no pressure, thanks for all your amazing work, and good luck with continued development!
Given the remote-first container-based world we're heading towards, decoupling UI (terminal emulator) from its "backend" (tmux, code-server) is a great design decision, which I think will ultimately define what the "next generation" of terminal emulators is. Imagine being able to open tabs directly on remote host, reconnect without losing state, etc, all while using native UI (so Cmd+T to open new tab, Cmd+F to search, etc). Productivity game changer, which currently only the iTerm2 users can fully enjoy.
Ptyxis (putting its backend in running containers), WezTerm (native handling of ssh sessions) and VSCode's terminal (starting a proprietary code-server binary and connecting to its TCP port) have reached some of this functionality, but in their design they need some out-of-band mechanisms to handle the connections, ultimately limiting the scenarios they can handle.
Meanwhile tmux -CC [0] and ht [1] are sending both their control channel and data channel over the opened terminal itself (in-band), making them flexible enough to support any configuration. Something complex like `ssh jumpbox -- ssh prod -- podman exec -it prod /bin/bash -- tmux -CC` should just work, as if everything was running on your local machine.
> Given the remote-first container-based world we're heading towards
You might want to expand on who "we" are here, because to me it seems to be a very small section of developers who want a "remote-first" experience and most (if not everyone) I speak with want software and tools that work local/offline-first, including our dear development environment.
If a tool requires a internet connection to work (so this "remote-first container-based world" you mention), I don't think either me nor many of the developers I know would even give it a try, because we need to be able to use our tools regardless if we have a working internet connection and/or the service-provider has downtime.
In fact, the only ones I know who are using a "remote-first" environment are developers who are forced to do so by their employer, and it's not a short-list of complaints about it when we meet over beers.
Not very small; most corporate non-Windows development uses containers, I'd bet on the ratio above 90%.
Not that developers want remote-first experience, but they face it. Containers in this regard behave like remote systems, even when run locally. A tool that helps juggle multiple remote contexts in a sane way may be very helpful. Say, tmux does a lot to help, but more is of course possible.
I'm not sure why they couldn't put that on the front page. Or at least on the front page of the documentation.
As it stands now, the front page is pretty useless if you never heard of ghostty. It's a ghost in a terminal? And there's tty in the name.... so is it a program to show ghosts in your terminal?
Absolutely agree, even if it was just the tagline on the first page of the documentation - "Ghostty is a fast, feature-rich, and cross-platform terminal emulator that uses platform-native UI and GPU acceleration." - underneath the "screenshot" and above the buttons on the homepage.
On mobile I thought this thing was some ascii art game for a hot minute before I tried clicking on the download link. Seems silly to admit that but there you have it.
I'd never heard of it before today, so yeah, my first thought was exactly what you postulated. "Cool? Maybe it's a tool for generating animations in the terminal?"
*edit - oh, no Windows support for now (my hands are tied at work).
How you've managed to communicate effectively with so many people who were using the beta version to craft the final release is beyond me!
I downloaded it last night, wrote one line of theme configuration (I picked the batman theme). and For me it just works fine and faster than any other terminal emulator I've ever worked with.
Also great with 'less' and 'tmux'.
I just switched to 'Ghostty'! so thank you.
Just downloaded it because of the animation on the site. How did you make that? it's cute.
The terminal is good. I don't have any issues with the stock mac os terminal app but I'll give yours a go for a bit. I can't tell exactly how, but it does feel faster - I'm not sure exactly how it is but it looks like it renders text faster.
I've been a beta tester from very early on. I came for the performance but stayed for the stability. I've only had a rare few crashes and all but one was a duplicate in the bug tracker.
I thought I needed search but as Mitchell put it, not a 1.0 feature. Ripgrep was always the answer.
Very happy to share the ghostty experience with the world!
I noticed a keybinding for write_scrollback_file (defaults to super+shift+j) which does this. So you can use `grep something <super+shift+j>` to search for things. It's a nice workaround until search is implemented.
Looks great. Switching from wezterm and kitty before that, and the native feel on macOS is indeed better than those emulated tabs. Also it is really quite zero config where to me it seems only one line of color theme is needed, together with setting TERM for ssh. (Default font happens to be mine.) Bundling so much color theme felt a bit waste of space, but it is nice that I don’t need to define them myself which also makes switching theme tedious. I also love to see it lands at 1.0 where features, documentation, and promotional materials seem polished already.
Chances are, it will become my default terminal in 2025!
The only nitpicking is the doc does not seem to have a search function. It would be great if I can search color or ssh and find the relevant page address that immediately, without navigating the tree or relying on external search engines.
I haven't heard of Ghostty before today. I'm reading the docs; but what are the main selling points?
Personally, I've used Terminal.app (for a bit), iTerm (for a long while), Konsole (in the past, on Linux) and more recently I standardized on Alacritty (macOS) and Kitty (Linux). GPU acceleration puts it in the same general bin as my current toolchain, but your comment alludes to it pulling in some concepts from iTerm, possibly -- is that the case?
How do you plan on sustaining the project? Do you have enough committers so you’ll not burn out? Will you need support? Is there a way to donate? I guess what I’m getting at is what is the long term sustainability work around this?
I think you've done an excellent job running the community for Ghostty and it is a prime example of how to do it right. From the Discord to Github repos you've been a class act all the way through and have pushed folks to be good, civil internet denizens. Much respect.
If anyone cares to search through Github, they will see loads and loads of Issues and PRs created by Mitchell in many of the related Open Source projects that Ghostty uses/references. From zig to kitty to supporting libraries, Mitchell has been trying to get the terminal community working together and have some sort of standards. A lot of them are like "X does this, Y does that, why are you doing it this way? Can we all do it this way?" and then having Ghostty follow the most reasonable solution (or supporting several!).
Just built this from source on Fedora and I have to say the experience even for building-from-source is flawless. Immediately replaced Kitty for me; it feels smoother and faster. Amazing work.
Alacritty on macOS and Linux user here (Windows Terminal on Windows due to easily different shells available, formerly used iTerm2 on macOS). I make up for lack of tabs with zellij locally (tmux remotely). Also allows me to relog or close/update Alacritty. I will give Ghostty a whirl but why no shout out to Alacritty? Which features am I missing out on?
Thanks. I'm interested in seeing examples (of Alacritty community not being nice). I'm aware of one now [1] but I have not been involved with the Alacritty community much. I read/write a few bug reports when the project was born; after that it has served me very well (compared to iTerm2 and Terminal.app). IIRC there was something with ligatures but I don't remember.
Thus far I have not been able to get my italic bold font (Cartograph CF) working on Ghostty, but this is a minor issue.
Yikes, that ramblerman who was all over that HN thread still posts today. The distance by which he missed the point is shocking. How does one get that far off without doing it deliberately?
I'm a fan of your work! I'm curious about how you decided to work on building a terminal for your next project among your other ideas. If you have time later, could you share your main motivation with us or link to an existing post if you already mentioned it elsewhere?
Shoutout to you sir for
Shouting out the other terminals. It’d be easy for someone of your fame and talent and history to ride the hype to the GOAT of all terminals. But you stayed humble. Props.
Thank you for building this! I’ve loved using this over the last two months or so and really appreciate the work you’ve put into it.
I’ve been a very happy iTerm2 user and support the dev on GitHub Sponsors (and I’ll continue to do that), but I love your commitment to making a fast, native app (and cross platform, no less) and really appreciate this very obvious labor of love that has also been really interesting to watch from afar as the development has progressed!
kitty has been my default for a few years now— highly superficial, but my designer side landed on kitty as it was one of the few that supports ligatures by default, specifically for the fira code font for my bias.
not only does it support them, but the default font provides them. although it splits the ligature in half if your cursor is in the middle of it which feels a little weird to me
what was the motivation to start this project? it got so many points here, hence I assume there must be something really useful about it - but I so far fail to see it.
Calling it libTerminal might be confusing. There are other terminal emulation libraries—for example libvterm[1], notably used to power the embedded terminal in Neovim.[2]
Looks really awesome. I'm going to sound like I don't belong in the hipster terminal club, but the reason I shied away from some of the other terminals is the lack of tabs, which looks like yours has when I did a quick Google question/search. (if wezterm and the like have them, I must have missed it or it wasn't obviously apparent in the settings how to achieve them).
I know everyone will say but tmux and/or native multiplexing bla, but I'm kind of old school and only do screen remotely if needed, and I just like a lot of terminal tabs in my workflow with a quick mod left/right arrow to navigate between (and if native multiplexing in Ghostty is simple and easy I'd probably do some of that too).
Perhaps this is why I've never left iterm2.
WezTerm has emulated tabs, not OS-native tabs. Some people will find that to be enough to suit their needs. Others will want 'real tabs'.
For example, I can select a tab in Ghostty, pull it out into its own window, and then stick it into another window. This doesn't work in WezTerm, nor can you drag them around to rearrange them (keyboard shortcuts allow this however).
I'm honestly not sure what a OS native tab is supposed to be, on Linux. And I do not see why "emulated tabs" of WezTerm couldn't just do the actions that you describe. The Firefox tabs do after all, and they do not look like they are built according to some (GTK?) tab standard (GTK.Notebook?). I'm pretty sure that X11 does not have native tabs, while some window manager do. Is there a linux standard I'm just not aware of that you are referencing here?
Maybe it does not matter, the difference in functionality should count. Just highlighting that this messaging might not be understandable for linux users. I guess you are talking about some macOS thingie that's irrelevant to us.
I also use tmux, but I love the native tabs of Konsole in KDE. I have Shift-Arrow configured to move between them, it is far more comfortable than the dual shortcuts needed by tmux, Ctrl-B to call tmux's attention then l (if I remember correctly) to get to the last tab.
Konsole also has easy resizing of text and supports images in the console, you might like it.
I second this. I wouldn't have been able to make tmux such an integral part of my daily workflow if it weren't for this binding that I came across in someone's starter tmux conf many years ago.
jj or qq can also be used. As I once started with GNU Screen ages ago, I like prefix to ctrl+a but it interferes with going to front of command in bash mode. For that, I use vi keybind mode in my shell (fish or bash) which feel very natural to me. For zellij, I use default prefix so that it does not interfere with remote (yes I am aware remote requires double prefix; I prefer using tmux remotely and zellij locally).
How are you configuring jj without interfering with using the key regularly? I used to have jj bound to ESC in VIM, loved it, but I've since trimmed back my config as I SSH into to many foreign systems to depend upon nonstandard behaviour. But I'd love to have it in Tmux.
The obvious `set -g prefix jj` throws an error that the key j is a bad key. Various experimenting with bind and unbind have not resulted in success, and I can't seem to find an example .tmux.conf with that config to copy.
As a longtime KDE fan I've used Konsole for years and it is seriously underrated. However, after I discovered Kitty a few months ago, I fully converted and cannot be happier. You can also move between tabs using keyboard shortcuts as well as move the tabs themselves. What I like about Kitty more is its straightforward text file based approach for configuration and absence of any gui elements. Konsole's GUI and configuration menus are overwhelming and I think are in need of a deep redesign.
I agree with the philosophy of no tabs, but I simultaneously live in an intermediate desktop environment situation where I don't have access to the "right" solution yet. So I'm happy to have terminal emulator tabs, browser tabs and IDE tabs for lack of a more integrated solution.
Wezterm has tabs right out of the box and they are fully customizable, though I prefer tmux since I prefer to not have my data extinguished if I accidentally close the terminal :D
WezTerm shines in ease and breadth of configurability due to using lua, so it's simple to have the theme change between light/dark depending on host OS theme.
Interestingly there’s another comment ITT complaining that they need to use a programming language for configuring wezterm :)
As a wezterm user I’ll admit that configuring it was mildly annoying to start, but ended up feeling like an accomplishment. A few years in, it’s just another annoying program I have to re-remember how to use when i update twice a year.
I don't know Lua either but c'mon now. It's just some config. Copy and paste from the docs and tweak the values as needed. If you really want to write something funky in there, ChatGPT knows Lua. It's not that bad.
Yeah. I could get by with the default Linux terminal and tmux really. Tmux is just the best. Second to vim it’s the single most useful thing I’ve ever used.
Ghostty got a lot of hype (I cover this in my reflection below), but I want to make sure I call out that there is a good group of EXCELLENT terminals out there, and I'm not claiming Ghostty is strictly better than any of them. Ghostty has different design goals and tradeoffs and if it's right for you great, but if not, you have so many good choices.
Shout out to Kitty, WezTerm, Foot in particular. iTerm2 gets some hate for being relatively slow but nothing comes close to touching it in terms of feature count. Rio is a super cool newer terminal, too. The world of terminals is great.
I’ve posted a personal reflection here, which has a bit more history on why I started this, what’s next, and some of the takeaways from the past two years. https://mitchellh.com/writing/ghostty-1-0-reflection