--- title: "Self-host VoiceStudio" description: "Run VoiceStudio on an exe.dev VM, no GPU needed: text to speech, voice cloning and transcription in your browser, an OpenAI-style API, and MCP." url: https://exe.dev/deploy/voicestudio category: "AI apps" updated: 2026-10-30 --- # Self-host VoiceStudio Your own text-to-speech and voice studio, on CPU, behind exe.dev login. - **Software:** VoiceStudio (https://voicestudio.sh) - **Source:** https://github.com/debpalash/VoiceStudio - **License:** AGPL-3.0 (each voice model has its own licence) - **Time:** 20 minutes - **Example VM name:** `my-voicestudio` (yours will differ; on the VM, `hostname` prints it) - **Machine-readable:** https://exe.dev/deploy/voicestudio/manifest.json ## What you get - VoiceStudio at the VM's private URL, locked with an API key kept in ~/voicestudio/.env. - Speech and transcription on the VM's CPU, checked with a short test clip. - Voices, projects, settings and models on the VM's disk, kept across restarts and updates. - Starts again after a reboot; nightly backups, with one checked. Running on the VM: - **VoiceStudio 0.5**: The official image, ghcr.io/debpalash/voicestudio, in Docker Compose, on CPU, port 8000, data in `~/voicestudio/data` - **Voice models**: Downloaded from Hugging Face when you pick them, kept beside your data, data in `~/voicestudio/data/huggingface` - **Nightly backup**: A systemd timer; voices, projects and settings, without the models, data in `~/voicestudio-backups` How it is reached: - https://my-voicestudio.exe.xyz/ (private): The studio, after exe.dev login and VoiceStudio's API key - https://my-voicestudio.exe.xyz/v1/ (private): OpenAI-style speech and transcription API, with a VM token and the API key - https://my-voicestudio.exe.xyz/mcp/ (private): MCP server for agents, with the same two credentials ## Deploy with Shelley Open https://exe.dev/new?prompt=Set%20up%20VoiceStudio%20on%20this%20VM%20by%20following%20the%20exe.dev%20recipe%20at%20https%3A%2F%2Fexe.dev%2Fdeploy%2Fvoicestudio.md%20from%20top%20to%20bottom.%0A%0AThe%20recipe%27s%20example%20VM%20name%20is%20my-voicestudio.%20This%20VM%27s%20name%20is%20the%20output%20of%20%60hostname%60%3A%20use%20it%20wherever%20the%20recipe%20writes%20my-voicestudio.%20Skip%20the%20step%20that%20creates%20the%20VM%3B%20you%20are%20on%20it.%0A%0AThis%20recipe%20needs%20a%2060GB%20disk%2C%20and%20a%20new%20VM%27s%20may%20be%20smaller.%20Before%20you%20install%20anything%2C%20run%20%60df%20-h%20%2F%60.%20If%20its%20size%20is%20under%2054G%2C%20stop%20and%20ask%20me%20to%20grow%20the%20disk%2C%20either%20with%20this%20link%3A%20https%3A%2F%2Fexe.dev%2Fsuggest%3Fcommand%3Dresize%2520my-voicestudio%2520--disk%253D60GB%20or%2C%20on%20my%20computer%2C%20with%20%60ssh%20exe.dev%20resize%20my-voicestudio%20--disk%3D60GB%60.%20When%20I%20say%20it%27s%20done%2C%20run%20%60sudo%20resize2fs%20%22%24%28findmnt%20-no%20SOURCE%20%2F%29%22%60%2C%20check%20%60df%20-h%20%2F%60%20again%2C%20and%20carry%20on.%0A%0AYou%20are%20done%20when%20all%20of%20these%20are%20true.%20Check%20each%20one%20yourself%3A%0A-%20VoiceStudio%20at%20the%20VM%27s%20private%20URL%2C%20locked%20with%20an%20API%20key%20kept%20in%20~%2Fvoicestudio%2F.env.%0A-%20Speech%20and%20transcription%20on%20the%20VM%27s%20CPU%2C%20checked%20with%20a%20short%20test%20clip.%0A-%20Voices%2C%20projects%2C%20settings%20and%20models%20on%20the%20VM%27s%20disk%2C%20kept%20across%20restarts%20and%20updates.%0A-%20Starts%20again%20after%20a%20reboot%3B%20nightly%20backups%2C%20with%20one%20checked.%0A%0ARead%20the%20whole%20recipe%2C%20including%20the%20closed%20details.%20For%20initial%20setup%2C%0Afollow%20%22Install%20by%20hand%22%2C%20skipping%20VM%20creation%20and%20SSH%3A%20room%20on%20the%20disk%2C%0ADocker%2C%20Start%20VoiceStudio%2C%20Check%20speech%2C%20the%20backup%20timer%20and%20the%20backup%0Acheck.%20The%20walkthrough%20at%20the%20top%20is%20for%20the%20owner%20afterwards.%20Do%20not%20run%0ARestore%20or%20Update.%0AWork%20as%20exedev%20in%20~%2Fvoicestudio.%20Never%20overwrite%20an%20existing%20.env.%0ADo%20not%20accept%20any%20model%20licence%20for%20the%20owner%2C%20and%20do%20not%20run%20the%0Afirst-run%20wizard%20in%20a%20browser%3A%20the%20owner%20does%20both.%20Do%20not%20share%20the%20VM%0Aor%20make%20it%20public.%0AVerify%20docker.service%20is%20enabled%20and%20the%20container%20uses%20unless-stopped%3B%0Ado%20not%20reboot.%20Finish%20with%20the%20studio%20URL%20and%20the%20command%20from%20%22Open%20the%0Astudio%22%20that%20prints%20the%20API%20key%3B%20don%27t%20paste%20the%20key%20into%20the%20chat.%0A%0AFinish%20with%20the%20URLs%20I%20should%20open.%0A to create a VM whose agent, Shelley, follows this recipe. From a terminal: ```sh curl -fsSL https://exe.dev/deploy/voicestudio/prompt.txt | ssh exe.dev new --disk=60GB --prompt=/dev/stdin ``` It needs a 60 GB disk. The terminal command makes the VM that size; after the button, Shelley checks and, if the disk is smaller, gives you a link to grow it. Or follow the steps below by hand. VoiceStudio is an open-source voice studio: type a script and hear it read aloud, design a voice from a description, clone one from a short clip, transcribe recordings, dub videos, and build audiobooks, in hundreds of languages. It runs its models on its own machine, so your scripts and recordings never go to a speech service. Here that machine is an exe.dev VM. There's no GPU, so VoiceStudio runs on the CPU: slower, but always on, private behind exe.dev login, and reachable from your browser, your apps through an OpenAI-style API, and your agents through MCP. Your voices and projects stay on the VM's disk, backed up every night. ## Open the studio VoiceStudio asks for an API key, its admin password. Shelley made one when it installed VoiceStudio. On your computer, print it: ```bash ssh my-voicestudio.exe.xyz sed -n s/^OMNIVOICE_API_KEY=//p voicestudio/.env ``` Open `https://my-voicestudio.exe.xyz/`, sign in to exe.dev if asked, and paste the key into **API key**. VoiceStudio keeps you signed in for up to eight hours, and until it restarts; then it asks for the key again. A password manager helps. The first time, a short setup follows: 1. **System check** finds no GPU and says everything runs on the CPU. Continue. 2. **Model packs**: choose **Fast** and **Install Fast pack** (2.7 GB, about two minutes). When every model says *installed*, press **Use Fast pack**, then continue. 3. **Privacy**: choose **No thanks**. The VM is set up never to send usage statistics anyway. 4. **Enter studio**. Skip *Try Dictation*; it is meant for the desktop app. The first time you synthesize with the default model, VoiceStudio asks you to accept its licence. Read it: the OmniVoice weights are for personal and research use only. If that doesn't fit, close it and [switch to KittenTTS](#choose-an-engine).
VoiceStudio's Design screen in the browser: a voice description box, starting points such as Narrator and Storyteller, a script, and a Synthesize audio button
Designing a voice in the studio, served from the VM on its CPU.
## Make some speech - **Design**: pick a starting point such as *Narrator*, or describe a voice ("a warm elderly British storyteller"), edit the script, and press **Synthesize audio**. **Save as voice profile** keeps the voice. - **Voice cloning**: add a recording with 5-15 seconds of clean speech, check the transcript, and save it as a voice. Only clone voices you have permission to use. - **Gallery** has hundreds of ready-made voices. **Audiobook**, **Stories** and **Dubbing** build longer pieces from them. - **Transcribe** turns uploaded recordings into text. Your takes are kept under **Recents** in the sidebar. Generated audio carries an invisible watermark marking it as AI-made; it's on by default, under **Settings → Privacy**. ## Choose an engine VoiceStudio has many speech engines; on a CPU, two matter. We measured them on a default VM (2 vCPUs, 8 GB): | Engine | Speed on 2 vCPUs | Voices | Model licence | | --- | --- | --- | --- | | VoiceStudio (OmniVoice), the default | about 10 s of work per second of speech | cloning, design, 600+ languages | CC-BY-NC: personal and research use | | KittenTTS | about 3x faster than real time | 8 preset voices, English | Apache-2.0 | Shelley installed KittenTTS (80 MB) while setting up. To make it the default, open **Settings → Models → TTS** and press **Select** on KittenTTS. Apps can also pick per request with `"model": "kittentts"`. For transcription, setup turned on Whisper base, which is quick (2 seconds for a short sentence) but not the most accurate. For better results, **Download** a larger model under **Settings → Models → ASR** and select it: Whisper large-v3 Turbo (1.6 GB) took 20 seconds for a 7-second clip. For faster speech, give the VM more CPUs. On your computer: ```bash ssh exe.dev resize my-voicestudio --cpu=4 --memory=16GB ``` ## Use it from apps VoiceStudio answers OpenAI's audio API: `POST /v1/audio/speech` and `POST /v1/audio/transcriptions`. The voice can be an OpenAI name like `alloy` (the engine's default voice) or one of your saved voices, by name or id; `GET /v1/audio/voices` lists them. On the VM itself, send the API key: ```bash key=$(sed -n 's/^OMNIVOICE_API_KEY=//p' ~/voicestudio/.env) curl -fsS http://localhost:8000/v1/audio/speech \ -H "Authorization: Bearer $key" -H 'Content-Type: application/json' \ -d '{"model":"kittentts","voice":"alloy","input":"Hello from my own voice server.","response_format":"wav"}' \ -o hello.wav ``` From anywhere else, the request first has to get past exe.dev's login. Make a token for this VM's [HTTPS proxy](/docs/https-tokens-for-vms), on your computer: ```bash ssh exe.dev ssh-key generate-api-key --vm=my-voicestudio --label=voicestudio ``` Send it as `X-Exedev-Authorization`, beside the VoiceStudio key. The proxy checks and removes the first; VoiceStudio checks the second: ```bash curl -fsS https://my-voicestudio.exe.xyz/v1/audio/speech \ -H "X-Exedev-Authorization: Bearer $EXE_TOKEN" \ -H "Authorization: Bearer $VOICESTUDIO_KEY" \ -H 'Content-Type: application/json' \ -d '{"model":"kittentts","voice":"alloy","input":"Hello from my own voice server.","response_format":"wav"}' \ -o hello.wav ``` In OpenAI's SDKs, set the base URL to `https://my-voicestudio.exe.xyz/v1`, the API key to the VoiceStudio key, and add the `X-Exedev-Authorization` header as a default header. ## Give your agents a voice VoiceStudio has an MCP server at `/mcp/` with tools to speak, clone and design voices, transcribe, and list voices. Claude Code on the VM connects over localhost; on the VM: ```bash key=$(sed -n 's/^OMNIVOICE_API_KEY=//p' ~/voicestudio/.env) claude mcp add --transport http voicestudio http://localhost:8000/mcp/ \ --header "Authorization: Bearer $key" claude mcp list ``` Expect `voicestudio: ... ✔ Connected`. An agent elsewhere uses `https://my-voicestudio.exe.xyz/mcp/` with both headers from [Use it from apps](#use-it-from-apps). The [MCP guide](https://github.com/debpalash/VoiceStudio/blob/main/docs/mcp.md) shows how to give each agent its own voice. ## Keep your voices Backups run nightly into `~/voicestudio-backups`: your voices, projects, history, settings, and the API key, but not the models, which download again. VoiceStudio stops for a few seconds while it's copied, so you may have to sign in again in the morning. Keep a copy **off the VM**. On your computer: ```bash scp -r my-voicestudio.exe.xyz:voicestudio-backups . ``` The backups hold your voice samples and the API key, so keep them private. To restore one or update VoiceStudio, see below, or ask Shelley. Something stuck? On the VM, try `cd ~/voicestudio && docker compose logs --tail=100`. ## The nuts and bolts Happy clicking Deploy? You can stop here. These are the same setup steps Shelley follows, plus the maintenance commands for later.
Install by hand ### Create a VM The VoiceStudio image is big (about 17 GB on disk, with PyTorch and CUDA libraries inside), so give the VM a 60 GB disk. You're billed for the disk you use, not its size. On your computer (pick another name if `my-voicestudio` is taken): ```bash ssh exe.dev new --name=my-voicestudio --disk=60GB ssh my-voicestudio.exe.xyz ``` All remaining installation commands run **on the VM**, as `exedev`. Commands use `hostname`, so your VM can have any name. ### Room on the disk Check the disk: ```bash df -h / ``` If **Size** is under 54G, the VM was made with a smaller disk. Grow it with [this link](https://exe.dev/suggest?command=resize%20my-voicestudio%20--disk%3D60GB) (with your VM's name), or on your computer with `ssh exe.dev resize my-voicestudio --disk=60GB`. Then, back on the VM, grow the filesystem to match: ```bash sudo resize2fs "$(findmnt -no SOURCE /)" df -h / ``` Expect a size of about 59G. ### Get Docker ready Enable Docker at boot, or exe.dev's on-demand Docker won't restart VoiceStudio after a reboot. Install `jq` if needed: ```bash sudo systemctl enable --now docker.service systemctl is-enabled docker.service command -v jq >/dev/null || sudo apt-get install -y jq ``` The check should print `enabled`. ### Start VoiceStudio This writes `.env` **once**, with a random API key, then a Compose file. The image listens on port 3900; Compose publishes it on port 8000, which exe.dev's proxy serves at `https://.exe.xyz/`. Everything VoiceStudio keeps, including the models it downloads, goes in `~/voicestudio/data`. ```bash mkdir -p ~/voicestudio/data cd ~/voicestudio if [ ! -e .env ]; then touch .env && chmod 600 .env echo "OMNIVOICE_API_KEY=$(openssl rand -hex 32)" >> .env fi cat > compose.yaml </dev/null do sleep 5; done done ``` Expect `install_started` twice. VoiceStudio's default transcription model is a 2.9 GB Whisper it hasn't downloaded, so make Whisper base the active one: ```bash auth="Authorization: Bearer $(sed -n 's/^OMNIVOICE_API_KEY=//p' ~/voicestudio/.env)" curl -fsS -X POST http://localhost:8000/engines/select -H "$auth" \ -H 'Content-Type: application/json' \ -d '{"family":"asr","backend_id":"faster-whisper","model_id":"Systran/faster-whisper-base"}'; echo ``` Expect `"active":"faster-whisper"`. Now have KittenTTS say a sentence, and Whisper write it back down: ```bash auth="Authorization: Bearer $(sed -n 's/^OMNIVOICE_API_KEY=//p' ~/voicestudio/.env)" curl -fsS http://localhost:8000/v1/audio/speech -H "$auth" \ -H 'Content-Type: application/json' \ -d '{"model":"kittentts","voice":"alloy","input":"The quick brown fox jumps over the lazy dog.","response_format":"wav"}' \ -o /tmp/hello.wav file /tmp/hello.wav curl -fsS http://localhost:8000/v1/audio/transcriptions -H "$auth" \ -F file=@/tmp/hello.wav -F model=whisper-1; echo rm /tmp/hello.wav ``` Expect `RIFF (little-endian) data, WAVE audio, ... mono 24000 Hz`, then `{"text":"The quick brown fox jumps over the lazy dog."}`. The default voice model, OmniVoice, waits for the owner to accept its licence in the studio. ### Set up nightly backups The backup stops VoiceStudio for a few seconds, so its database is copied in a consistent state. It runs as root because the container writes `data` as root, and skips `data/huggingface`, the downloaded models. ```bash cat > ~/voicestudio/backup.sh <<'EOF' #!/usr/bin/env bash # Back up VoiceStudio: data (voices, projects, history, settings) without # the downloadable models, .env (the API key), and the Compose file. # Run with sudo. set -euo pipefail umask 077 cd "$(dirname "$(readlink -f "$0")")" owner=$(stat -c %U:%G .) dest="$(dirname "$PWD")/voicestudio-backups" mkdir -p "$dest" out="$dest/voicestudio-$(date -u +%Y%m%d-%H%M%S).tar.gz" # A short stop keeps the SQLite database consistent. docker compose stop voicestudio trap 'docker compose start voicestudio' EXIT tar -czf "$out.part" --exclude=data/huggingface .env compose.yaml backup.sh data docker compose start voicestudio trap - EXIT mv "$out.part" "$out" chown -R "$owner" "$dest" # Keep the 14 newest backups. ls -1t "$dest"/voicestudio-*.tar.gz | tail -n +15 | xargs -r rm -- echo "$out" EOF chmod +x ~/voicestudio/backup.sh ``` Schedule it for 03:30 UTC (or so), then take a first backup: ```bash sudo tee /etc/systemd/system/voicestudio-backup.service >/dev/null </dev/null <<'EOF' [Unit] Description=Back up VoiceStudio every night [Timer] OnCalendar=*-*-* 03:30:00 RandomizedDelaySec=30m Persistent=true [Install] WantedBy=timers.target EOF sudo systemctl daemon-reload sudo systemctl enable --now voicestudio-backup.timer sudo systemctl start voicestudio-backup.service systemctl list-timers voicestudio-backup.timer ls -lh ~/voicestudio-backups until curl -fs http://localhost:8000/health; do sleep 5; done; echo ``` Check that the timer has a next run, the directory has a tarball of a few hundred KB (more once you have voices), and VoiceStudio is back. For failures: `journalctl -u voicestudio-backup.service`. ### Test a backup Unpack the newest backup into a **scratch directory** and check its database, without touching the running copy: ```bash latest=$(ls -1t ~/voicestudio-backups/voicestudio-*.tar.gz | head -n 1) mkdir /tmp/voicestudio-restore-test sudo tar -xzf "$latest" -C /tmp/voicestudio-restore-test sudo ls -A /tmp/voicestudio-restore-test /tmp/voicestudio-restore-test/data sudo python3 -c ' import sqlite3 db = sqlite3.connect("file:/tmp/voicestudio-restore-test/data/omnivoice.db?mode=ro", uri=True) print(db.execute("pragma integrity_check").fetchone()[0]) print("voices", db.execute("select count(*) from voice_profiles").fetchone()[0]) ' sudo rm -rf /tmp/voicestudio-restore-test ``` Expect `.env`, `compose.yaml`, `backup.sh` and `data` (with `omnivoice.db` and no `huggingface`), then `ok` and the number of saved voices. Installation done. Continue with [Open the studio](#open-the-studio).
Back up and restore On the VM, `sudo ~/voicestudio/backup.sh` makes an extra backup and prints its path. Copy backups off the VM as shown in [Keep your voices](#keep-your-voices). ### Restore On a new VM, follow [Install](#create-the-vm) up to and including [Get Docker ready](#get-docker-ready). Create `~/voicestudio-backups` there, then copy your backup over from your computer: ```bash scp voicestudio-20261030-033012.tar.gz my-voicestudio.exe.xyz:voicestudio-backups/ ``` Use your backup's actual filename in this and the commands below. All remaining commands run **on the VM**. If VoiceStudio already runs there, stop it and keep its directory aside first: ```bash cd ~/voicestudio && docker compose down sudo mv ~/voicestudio ~/voicestudio.before-restore ``` Unpack the backup, point MCP at this VM's name, and start it: ```bash mkdir ~/voicestudio sudo tar -xzf ~/voicestudio-backups/voicestudio-20261030-033012.tar.gz -C ~/voicestudio cd ~/voicestudio sed -i "s/^\( *OMNIVOICE_MCP_ALLOWED_HOSTS:\).*/\1 $(hostname).exe.xyz/" compose.yaml docker compose up -d until curl -fs http://localhost:8000/health; do sleep 5; done; echo ``` Sign in with the API key from the backup and check your voices are there. The models aren't in the backup: download KittenTTS and Whisper base again with the first block in [Check speech](#check-speech), and the rest from **Settings → Models**. On a new VM, recreate the backup timer from [Install](#set-up-nightly-backups). Keep `~/voicestudio.before-restore` until you're happy everything is back.
Update VoiceStudio The `0.5` image tag follows the newest 0.5.x release. On the VM, back up, then pull and restart: ```bash cd ~/voicestudio sudo ./backup.sh docker compose pull docker compose up -d until curl -fs http://localhost:8000/health; do sleep 5; done; echo docker image prune -f df -h / ``` The health check shows the new version. The prune frees the old image's disk space; until then, both images are on disk, which is why the VM has 60 GB. Sign in to the studio again after the restart. For a new minor version, such as 0.6, read the [release notes](https://github.com/debpalash/VoiceStudio/releases) first, then change the `image:` tag in `compose.yaml` and run the same commands. VoiceStudio upgrades its database on start and keeps a copy of the old one in `data`; to roll back fully, [restore the backup](#restore).
Something not working? On the VM: ```bash cd ~/voicestudio docker compose ps docker compose logs --tail=100 voicestudio ``` - **502 right after starting or updating:** VoiceStudio takes a minute to load. Wait for `curl -fsS http://localhost:8000/health`. - **The studio asks for the API key again:** sessions end after eight hours and whenever VoiceStudio restarts, including during the nightly backup. Paste the key from `~/voicestudio/.env`. - **"This model can't be used until you accept its licence":** open the licence dialog when you synthesize and accept it, or select another engine such as KittenTTS under **Settings → Models → TTS**. - **"No speech-to-text model is installed":** under **Settings → Models → ASR**, select a Whisper model marked *installed*, or download one. - **A generation times out:** CPU synthesis of long scripts is slow. Split the script, use KittenTTS, or give the VM more CPUs with `ssh exe.dev resize`. - **MCP says `Invalid Host header`:** the VM was renamed or restored on another VM. Set `OMNIVOICE_MCP_ALLOWED_HOSTS` in `compose.yaml` to the current `.exe.xyz` and run `docker compose up -d`. - **Down after a reboot:** run `sudo systemctl enable --now docker.service`. - **Disk full:** check `df -h /` and `sudo du -sh ~/voicestudio/data/huggingface`. Run `docker image prune -f`, or grow the disk with `ssh exe.dev resize`.
## Frequently asked questions ### Does VoiceStudio need a GPU? No. Everything runs on the CPU, just slower. On the default 2-vCPU VM, KittenTTS speaks faster than real time, while the default VoiceStudio (OmniVoice) model, which clones and designs voices, takes about ten seconds per second of speech. More CPUs help: see [Choose an engine](#choose-an-engine). ### Can I use the audio commercially? It depends on the model, not the app. VoiceStudio itself is AGPL-3.0, but the default OmniVoice weights are licensed CC-BY-NC, for personal and research use, and VoiceStudio asks you to accept that before first use. KittenTTS is Apache-2.0. **Settings → Models** links each model's licence; check it before you publish anything. ### Can I clone my own voice? Yes. Add 5-15 seconds of clean speech under **Voice cloning**, save it as a voice, and use it in the studio, the API, or MCP. Only clone voices you have permission to use. ### Who can use my VoiceStudio? Only people you [share the VM](/docs/sharing) with get past exe.dev's login, and they also need the API key, which is VoiceStudio's admin password. Everyone with the key sees the same voices and history, so share it only with people you trust. ### Can my apps use it instead of a hosted speech API? Yes. VoiceStudio speaks OpenAI's audio API (`/v1/audio/speech`, `/v1/audio/transcriptions`), so most SDKs only need a new base URL and two headers. See [Use it from apps](#use-it-from-apps). ### How much does it cost? VoiceStudio is free. You pay for your [VM plan](https://exe.dev/pricing); the image and a few models use about 25 GB of disk. ## Related - [Self-host Open WebUI](https://exe.dev/deploy/open-webui.md): A private ChatGPT for you and your team. No API keys. - [Self-host Hermes Agent](https://exe.dev/deploy/hermes-agent.md): An agent that remembers you, on a computer that never sleeps. - [Run Claude Code on a remote server](https://exe.dev/deploy/claude-code.md): A computer for Claude. A breather for your laptop. - [Self-host n8n](https://exe.dev/deploy/n8n-2.md): Connect your apps. Let the boring bits run themselves. - [Host OpenClaw](https://exe.dev/deploy/openclaw.md): An AI sidekick with its own computer. Your laptop can take a nap. - [VS Code in your browser, with code-server](https://exe.dev/deploy/code-server.md): Your editor in a tab. Your projects right where you left them.