Asciinema alternatives: what you lose and gain with each one
_People look for an asciinema alternative because of one missing thing, not because asciinema is bad. Here is each replacement sorted by the thing you are actually after (video, GIF, SVG, a repeatable script or a narrated Short), with output format, OS, license and last-commit date checked on 3 November 2026._
I build ReelMint, a video studio for developers making vertical Shorts, so I have a stake in terminal demos. I also use asciinema, and I am not trying to talk you out of it. If you searched for an asciinema alternative, you almost certainly hit one specific wall: you needed a video or a GIF and got a text recording. This post sorts the alternatives by which wall you hit. For a wider comparison of recording tools in general, our roundup of terminal recording tools for CLI demos covers the field; this one is narrower and about replacing asciinema specifically.
Everything below with a date was checked against the tool's GitHub repository or README on November 3, 2026. Where I could not confirm something, I say so instead of guessing.
What asciinema does well, and the three things people replace it for
asciinema is a terminal session recorder and streamer. Its README describes recording to a lightweight text-based asciicast file, for example asciinema rec demo.cast, and it can also live-stream a session. As of November 3, 2026 the repository is active (last commit October 27, 2026), written in Rust, and licensed GPL-3.0. It runs on GNU/Linux, macOS and FreeBSD, and the README says Windows is not supported.
That text-first design is the reason it plays back crisply at any size. It is also the source of the three common reasons people leave:
- No video output. The README contrasts asciinema with screen recorders that produce heavyweight .mp4 and .mov files. It records .cast, not video. Neither the README nor the CLI docs page I read mention an MP4 or GIF export.
- No editing step. The docs pages I read describe recording, streaming and playback, not editing. If a take goes wrong, the path I could verify is to record again.
- Dependence on the player. A .cast file plays in the asciinema web player, which the docs say you can use on any HTML page. That works for docs and READMEs that allow JavaScript. It does not help on a platform that wants a video file, which includes YouTube, LinkedIn and most social feeds.
The alternatives, sorted by the job you need done
Pick the job first. The tools below fall into four groups, and most of the comparison is about which group you are in.
Job one: keep asciinema, add a GIF. agg is the asciinema project's own GIF generator. It converts asciicast v1, v2 and v3 recordings to animated GIF, using the gifski encoder, with built-in color themes, adjustable fonts, playback speed and an idle-time limit. It outputs GIF only. If your only complaint was that a README does not play a .cast, this is the smallest change that fixes it, and you keep your recording workflow.
Job two: a repeatable script that renders GIF, MP4 or WebM. VHS from Charm is the strongest answer to the no-video complaint. You write a .tape file (Type, Enter, Sleep, and Set for font and size), run vhs demo.tape, and get the output you declared. Its README lists GIF, MP4, WebM and a PNG frame sequence, and prebuilt binaries for Linux, macOS and Windows. It needs ttyd and ffmpeg on your PATH, or you can use its Docker image. Because the session is scripted rather than recorded live, the same tape gives the same result each time, which also makes it easy to regenerate in CI.
# demo.tape
Output demo.mp4
Set FontSize 28
Set Width 1200
Type "npm install -g yourtool"
Enter
Sleep 2s
# render it
vhs demo.tape
Job three: record a real session and get a GIF, with themes. terminalizer records a session to a YAML file, then renders it as an animated GIF or a web player, on Linux, Windows and macOS. The caveat is maintenance: its last commit was August 29, 2024, and its README says it expects Node.js v4 to v16 and that GIF compression is not implemented. That is not a verdict that it is broken, only what I could verify. t-rec takes a different approach and screenshots the terminal window (or another window) at 4 to 15 fps, producing a GIF and optionally an MP4. It runs on macOS, Linux, NetBSD and, in preview, Windows, and depends on ImageMagick. Its last commit to the default branch was July 8, 2026, a preview release.
Job four: an SVG animation. termtosvg records a session as an SVG animation. The repository is archived on GitHub, with its last commit on June 16, 2020, so treat it as finished. It suits a sharp, small README animation if you accept an unmaintained tool; otherwise pick agg.
Plain screen recording. OBS or QuickTime pointed at a terminal window gives you any video file, with no new tool to learn. The trade-off is that it is one take of pixels at your window's size and aspect ratio, so a typo means a re-record.
Terminal built-ins. I looked for a recorder in iTerm2 and Warp worth listing here, and could not confirm either well enough to compare them fairly, so I am leaving them out rather than guess.
Comparison: format, OS, license and last commit
One line per tool. Dates are the latest commit on the default branch as of November 3, 2026.
- asciinema: output: .cast text recording (no video); OS: Linux, macOS, FreeBSD; license: GPL-3.0; last commit: 2026-10-27.
- agg: output: GIF from asciicast; OS: I could not confirm a list (installs via Cargo or Docker); license: GPL-3.0; last commit: 2026-10-27.
- VHS: output: GIF, MP4, WebM, PNG frames; OS: Linux, macOS, Windows; license: MIT; last commit: 2026-09-24.
- terminalizer: output: GIF or web player; OS: Linux, Windows, macOS; license: MIT; last commit: 2024-08-29.
- t-rec: output: GIF, optional MP4; OS: macOS, Linux, NetBSD, Windows (preview); license: GPL-3.0; last commit: 2026-07-08.
- termtosvg: output: SVG animation; OS: not checked; license: BSD-3-Clause; repository archived, last commit: 2020-06-16.
- OBS / QuickTime: output: any video file; OS: varies by app; license: varies by app; no repository checked.
- ReelMint terminal scene: output: a vertical video project published to YouTube Shorts; web app plus an MCP server; hosted product, free tier, bring your own AI keys.
Which asciinema alternative should you pick?
- You want a GIF in a README and already use asciinema: add agg.
- You want MP4 or WebM, or the demo regenerated on every release: use VHS.
- You must capture a real session, including a TUI, and want a GIF: try t-rec or terminalizer, and check their maintenance dates above first.
- You want a small SVG animation and accept an archived tool: termtosvg.
- The destination is a phone-sized video with a voiceover: none of the recorders above reflow a desktop terminal into 9:16, so see the next section.
Is there an asciinema alternative with narration or captions?
Not among the recorders I checked: their documentation describes terminal output, not narration, and I found no captions feature in any of them. For a narrated terminal clip, the usual route is to record or script the terminal, then add voice and captions in a separate video editor.
That gap is the one ReelMint is built for. Instead of recording a shell, a terminal demo scene takes the command and the output as text, types them out on a 9:16 canvas, and sits alongside narration in the same editor. Because the command is text, fixing a renamed flag is an edit and a re-render. It also means nothing from your real shell, such as a stray key or hostname, can end up on screen. You can drive it from Claude Desktop, Claude Code or Cursor over an MCP server, and it publishes to YouTube Shorts. It runs on your own provider keys (BYOK) with a free tier.
When ReelMint is the wrong pick, and when to just keep asciinema
ReelMint does not record. If the point of your clip is proof that something really ran, such as a benchmark, a live deploy or an interactive TUI responding to keystrokes, you need a real recorder: asciinema, VHS or a screen capture. A typed scene can only show what you tell it to show.
If your demo lives in documentation, keep asciinema. A .cast file is small, text-based, copyable and plays at any size in the web player. For a README that needs to run on GitHub, where scripts do not execute, add agg for the GIF rather than switching tools. Move on only when the destination changes to somewhere that demands a video file.
Can asciinema export to mp4 or gif?
Not by itself, as far as the README and CLI docs I read show. asciinema records .cast files. The asciinema project publishes a separate tool, agg, that converts those recordings to animated GIF. For MP4, the closest direct option in this list is VHS, which renders from a script rather than from a .cast file.
What is the best asciinema alternative for a GIF in a README?
It depends on whether you already have recordings. If you do, agg turns an asciicast into a GIF without re-recording. If you are starting from scratch and want it regenerated on every release, VHS can output a GIF from a .tape script and is actively maintained (last commit September 24, 2026). For an animated, vertical, narrated version of the same demo, see how to turn a README into a demo video.
---
Originally published on reelmint.io.
Mentions
Asciinema alternatives: what you lose and gain with each one _People look for an asciinema alternative because of one missing thing, not because asciinema is bad. Here is each replacement sorted by the thing you are actually after (video, GIF, SVG, a repeatable script or a narrated Short), with output format, OS, license and last-commit date checked on 3 November 2026._ I build ReelMint, a video studio for developers making vertical Shorts, so I have a stake in terminal demos. I also use asciinema, and I am not trying to talk you out of it. If you searched for an asciinema alternative, you almost certainly hit one specific wall: you needed a video or a GIF and got a text recording. This post sorts the alternatives by which wall you hit. For a wider comparison of recording tools in general, our roundup of terminal recording tools for CLI demos covers the field; this one is narrower and about replacing asciinema specifically. Everything below with a date was checked against the tool's GitHub repository or README on November 3, 2026. Where I could not confirm something, I say so instead of guessing. What asciinema does well, and the three things people replace it for asciinema is a terminal session recorder and streamer. Its README describes recording to a lightweight text-based asciicast file, for example asciinema rec demo.cast, and it can also live-stream a session. As of November 3, 2026 the repository is active (last commit October 27, 2026), written in Rust, and licensed GPL-3.0. It runs on GNU/Linux, macOS and FreeBSD, and the README says Windows is not supported. That text-first design is the reason it plays back crisply at any size. It is also the source of the three common reasons people leave: - No video output. The README contrasts asciinema with screen recorders that produce heavyweight .mp4 and .mov files. It records .cast, not video. Neither the README nor the CLI docs page I read mention an MP4 or GIF export. - No editing step. The docs pages I read describe recording, streaming and playback, not editing. If a take goes wrong, the path I could verify is to record again. - Dependence on the player. A .cast file plays in the asciinema web player, which the docs say you can use on any HTML page. That works for docs and READMEs that allow JavaScript. It does not help on a platform that wants a video file, which includes YouTube, LinkedIn and most social feeds. The alternatives, sorted by the job you need done Pick the job first. The tools below fall into four groups, and most of the comparison is about which group you are in. Job one: keep asciinema, add a GIF. agg is the asciinema project's own GIF generator. It converts asciicast v1, v2 and v3 recordings to animated GIF, using the gifski encoder, with built-in color themes, adjustable fonts, playback speed and an idle-time limit. It outputs GIF only. If your only complaint was that a README does not play a .cast, this is the smallest change that fixes it, and you keep your recording workflow. Job two: a repeatable script that renders GIF, MP4 or WebM. VHS from Charm is the strongest answer to the no-video complaint. You write a .tape file (Type, Enter, Sleep, and Set for font and size), run vhs demo.tape, and get the output you declared. Its README lists GIF, MP4, WebM and a PNG frame sequence, and prebuilt binaries for Linux, macOS and Windows. It needs ttyd and ffmpeg on your PATH, or you can use its Docker image. Because the session is scripted rather than recorded live, the same tape gives the same result each time, which also makes it easy to regenerate in CI. # demo.tape Output demo.mp4 Set FontSize 28 Set Width 1200 Type "npm install -g yourtool" Enter Sleep 2s # render it vhs demo.tape Job three: record a real session and get a GIF, with themes. terminalizer records a session to a YAML file, then renders it as an animated GIF or a web player, on Linux, Windows and macOS. The caveat is maintenance: its last commit was August 29, 2024, and its README says it expects Node.js v4 to v16 and that GIF compression is not implemented. That is not a verdict that it is broken, only what I could verify. t-rec takes a different approach and screenshots the terminal window (or another window) at 4 to 15 fps, producing a GIF and optionally an MP4. It runs on macOS, Linux, NetBSD and, in preview, Windows, and depends on ImageMagick. Its last commit to the default branch was July 8, 2026, a preview release. Job four: an SVG animation. termtosvg records a session as an SVG animation. The repository is archived on GitHub, with its last commit on June 16, 2020, so treat it as finished. It suits a sharp, small README animation if you accept an unmaintained tool; otherwise pick agg. Plain screen recording. OBS or QuickTime pointed at a terminal window gives you any video file, with no new tool to learn. The trade-off is that it is one take of pixels at your window's size and aspect ratio, so a typo means a re-record. Terminal built-ins. I looked for a recorder in iTerm2 and Warp worth listing here, and could not confirm either well enough to compare them fairly, so I am leaving them out rather than guess. Comparison: format, OS, license and last commit One line per tool. Dates are the latest commit on the default branch as of November 3, 2026. - asciinema: output: .cast text recording (no video); OS: Linux, macOS, FreeBSD; license: GPL-3.0; last commit: 2026-10-27. - agg: output: GIF from asciicast; OS: I could not confirm a list (installs via Cargo or Docker); license: GPL-3.0; last commit: 2026-10-27. - VHS: output: GIF, MP4, WebM, PNG frames; OS: Linux, macOS, Windows; license: MIT; last commit: 2026-09-24. - terminalizer: output: GIF or web player; OS: Linux, Windows, macOS; license: MIT; last commit: 2024-08-29. - t-rec: output: GIF, optional MP4; OS: macOS, Linux, NetBSD, Windows (preview); license: GPL-3.0; last commit: 2026-07-08. - termtosvg: output: SVG animation; OS: not checked; license: BSD-3-Clause; repository archived, last commit: 2020-06-16. - OBS / QuickTime: output: any video file; OS: varies by app; license: varies by app; no repository checked. - ReelMint terminal scene: output: a vertical video project published to YouTube Shorts; web app plus an MCP server; hosted product, free tier, bring your own AI keys. Which asciinema alternative should you pick? - You want a GIF in a README and already use asciinema: add agg. - You want MP4 or WebM, or the demo regenerated on every release: use VHS. - You must capture a real session, including a TUI, and want a GIF: try t-rec or terminalizer, and check their maintenance dates above first. - You want a small SVG animation and accept an archived tool: termtosvg. - The destination is a phone-sized video with a voiceover: none of the recorders above reflow a desktop terminal into 9:16, so see the next section. Is there an asciinema alternative with narration or captions? Not among the recorders I checked: their documentation describes terminal output, not narration, and I found no captions feature in any of them. For a narrated terminal clip, the usual route is to record or script the terminal, then add voice and captions in a separate video editor. That gap is the one ReelMint is built for. Instead of recording a shell, a terminal demo scene takes the command and the output as text, types them out on a 9:16 canvas, and sits alongside narration in the same editor. Because the command is text, fixing a renamed flag is an edit and a re-render. It also means nothing from your real shell, such as a stray key or hostname, can end up on screen. You can drive it from Claude Desktop, Claude Code or Cursor over an MCP server, and it publishes to YouTube Shorts. It runs on your own provider keys (BYOK) with a free tier. When ReelMint is the wrong pick, and when to just keep asciinema ReelMint does not record. If the point of your clip is proof that something really ran, such as a benchmark, a live deploy or an interactive TUI responding to keystrokes, you need a real recorder: asciinema, VHS or a screen capture. A typed scene can only show what you tell it to show. If your demo lives in documentation, keep asciinema. A .cast file is small, text-based, copyable and plays at any size in the web player. For a README that needs to run on GitHub, where scripts do not execute, add agg for the GIF rather than switching tools. Move on only when the destination changes to somewhere that demands a video file. Can asciinema export to mp4 or gif? Not by itself, as far as the README and CLI docs I read show. asciinema records .cast files. The asciinema project publishes a separate tool, agg, that converts those recordings to animated GIF. For MP4, the closest direct option in this list is VHS, which renders from a script rather than from a .cast file. What is the best asciinema alternative for a GIF in a README? It depends on whether you already have recordings. If you do, agg turns an asciicast into a GIF without re-recording. If you are starting from scratch and want it regenerated on every release, VHS can output a GIF from a .tape script and is actively maintained (last commit September 24, 2026). For an animated, vertical, narrated version of the same demo, see how to turn a README into a demo video. --- Originally published on reelmint.io.