Keyboard and office sound effects you generate yourself

Typing is one of the most-searched sounds on free libraries and one of the most annoying to source, because you usually need something like exactly eleven seconds of typing that stops on the beat. Generating gets you the length you asked for, and the right keyboard: a membrane laptop, a heavy mechanical board and a 1960s typewriter sound completely different, and a search for "typing" mixes them together.

Listen: 10 samples from the keyboard & office collection

Every clip shows the exact prompt that produced it. The render script reads its prompts from the same file as this page. Copy one, change a word and generate your own.

  • Mechanical typing

    via audiocpp-stable-audio-3-sfx

    typing steadily on a mechanical keyboard with tactile switches, distinct clacking key presses, close mic, quiet office

  • Laptop typing

    via audiocpp-stable-audio-3-sfx

    typing on a laptop membrane keyboard, soft muted key taps at a steady pace, close mic, quiet room

  • Single key

    via audiocpp-stable-audio-3-sfx

    single mechanical keyboard key pressed and released, crisp click with a bottom-out thock, very close mic, dry

  • Enter key

    via audiocpp-stable-audio-3-sfx

    enter key struck hard on a mechanical keyboard, deep decisive clack, very close mic, dry

  • Typewriter

    via audiocpp-stable-audio-3-sfx

    vintage manual typewriter typing a line then the carriage return bell and lever, close mic, small office

  • Mouse click

    via audiocpp-stable-audio-3-sfx

    computer mouse clicked twice in quick succession, tiny sharp plastic clicks, very close mic, dry

  • Scroll wheel

    via audiocpp-stable-audio-3-sfx

    mouse scroll wheel turned through its detents, fine ratcheting ticks, very close mic, dry

  • Pen click

    via audiocpp-stable-audio-3-sfx

    retractable ballpoint pen clicked repeatedly, sharp plastic snaps, very close mic, quiet room

  • Frantic typing

    via audiocpp-stable-audio-3-sfx

    someone typing very fast and erratically on a mechanical keyboard, dense overlapping key presses, close mic

  • Desk ambience

    via audiocpp-stable-audio-3-sfx

    quiet open-plan office ambience, distant typing, a phone ringing far away and low ventilation hum, seamless loop

Prompts that work

These models respond to physical description. Name the material, the action and the space.

  • Mechanical board

    typing steadily on a mechanical keyboard with tactile switches, distinct clacking key presses, close mic, quiet office

    "Tactile switches" is a real spec with a real sound. It gets the click without the mushiness of a membrane board.

  • Laptop board

    typing on a laptop membrane keyboard, soft muted key taps at a steady pace, close mic, quiet room

    "Membrane" and "muted" together stop the model reaching for the louder mechanical recordings it has more of.

  • Pace as a control

    someone typing very fast and erratically on a mechanical keyboard, dense overlapping key presses, close mic

    Pace carries the character's state: steady sounds like work, erratic sounds like panic. Set it in the prompt instead of in the edit.

  • Single event

    single mechanical keyboard key pressed and released, crisp click with a bottom-out thock, very close mic, dry

    "Pressed and released" gets both halves of the key travel, which makes a UI click sound physical.

Making keyboard & office sounds that fit

Render the length of the shot

A cut needs typing that runs for the exact duration of a shot and stops cleanly. A library gives you fourteen seconds when you needed nine, so you fade it and it sounds faded. Set the render length to the shot length. Every prompt here carries an explicit duration for this reason.

Choose the keyboard

A hacker scene, an office comedy and a period drama need three different keyboards. Mechanical boards are bright, clicky and percussive, with an audible bottom-out. Membrane laptops are dull, fast and close to silent. Typewriters are loud and mechanical, and the carriage return is a beat of its own. Pick one and name it in the prompt; that does more for a scene than processing.

Add desk foley

A room with only typing in it sounds like a sound effect. Real desks have pens, mice, mugs, chairs and a ventilation hum, and the small items make the space sound inhabited. Generate two or three of them separately and lay them under the typing at a low level. The pen click and the scroll wheel in this collection are meant for this, and they work better as a bed than in the foreground.

Why generate locally

The models run on your machine, which changes the cost, the privacy and what you can automate.

Offline after download

Weights are fetched once. After that, generation, the library and the API all work with the network off. Everything lives in one SQLite file and one folder of WAVs under ~/.soundbox, and nothing is uploaded.

Your agent can use it

Soundbox runs an MCP server in the same process, with tools for generating and listing sounds. Generated files come back as absolute paths, so an agent can copy them into a project.

Licences stated up front

Every model's terms are shown before you download it, and non-commercial weights need an explicit confirmation. Soundbox is MIT. The licence of the model that made a sound governs what you may do with it.

Connect a coding agent
claude mcp add --transport http soundbox http://127.0.0.1:17494/mcp
Generate over HTTP
curl -X POST http://127.0.0.1:17494/api/generate \
  -H 'Content-Type: application/json' \
  -d '{"model":"audiocpp-stable-audio-3-sfx","prompt":"8-bit coin pickup blip","duration_s":1}'

Frequently asked questions

Can I get typing that matches a specific words-per-minute?
Not precisely. The models respond to pace words like "steady", "fast", "erratic" and "hunt and peck", and not to a number. That is usually enough, because a scene needs the character's state more than a measurement. If you need exact sync to what is on screen, generate a single key press and place it per keystroke. That gives you full control for one render.
Can I use these sounds in a commercial video or game?
It depends on the model that made them. Soundbox shows each model's licence before you download its weights. The clips on this page were rendered with Stable Audio 3.0 Small SFX, which uses the Stability AI Community License: commercial use is free below USD $1M annual revenue, and above that Stability requires an enterprise licence. Soundbox itself is MIT, but the audio is governed by the licence of the model that made it. Read that licence before you ship; this summary is not legal advice.
Does this work offline?
Yes, once the weights are on disk. Generation, the library and the HTTP API all run locally against one SQLite file and one folder of WAVs under ~/.soundbox. There is no account and no API key, and nothing is uploaded. The only step that uses the network is the initial model download.
What hardware does it need?
The clips here were rendered on a consumer NVIDIA GPU at roughly ten seconds of wall time per clip, through audio.cpp's quantised GGUF runtime, which needs little VRAM. CPU-only generation is supported, but expect minutes per clip instead of seconds: fine for a handful of sounds, slow for a whole library.
Can an agent generate these for me?
Yes. While the app is open, Soundbox runs an MCP server at http://127.0.0.1:17494/mcp with tools for generating, listing and mixing sounds. Every generated file comes back as an absolute path, so a coding agent can render the sound a scene needs and put it straight into the project. The several hundred clips across these pages were made this way.