How to Add Subtitles in Kdenlive

Creating subtitles manually can consume a significant part of your production schedule, especially when you are facing tight deadlines or managing multiple video projects. Kdenlive offers a powerful, free solution for video editing and subtitle creation, but its local speech-to-text features require additional models and dependencies.

Combining Kdenlive’s editing capabilities with automated transcription can reduce the amount of manual transcription and synchronization required, giving you an efficient workflow without replacing Kdenlive as your primary video editor.

Key Takeaways

Kdenlive is a free, open-source video editor that can import ASS, SRT, VTT, and SBV subtitle files and export subtitles as ASS or SRT.
Manual subtitle creation involves transcribing, timing, reviewing, and formatting each caption.
Kdenlive includes local speech recognition through VOSK and Whisper, but both require speech models and supporting dependencies.
Kdenlive’s documentation describes Whisper as generally more accurate than VOSK, although results vary by model, language, audio quality, and hardware.
An external transcription service can generate a timed subtitle file that you can import and refine in Kdenlive.
Kdenlive supports subtitle styling, multiple subtitle files and layers, and burned-in or embedded subtitles.
Sonix offers automated transcription with up to 99% accuracy for clear recordings, transcription in 54+ languages, and translation into 55+ languages.

Understanding Subtitles and Why They Matter

Before diving into Kdenlive’s subtitle tools, it is worth understanding why subtitles and captions matter for video content.

Accessibility support is one of their most important uses. Accurate, synchronized captions can help meet accessibility requirements for prerecorded video. Accessibility captions typically include meaningful non-dialogue audio, such as music, sound effects, and speaker identification. Adding dialogue-only subtitles does not by itself guarantee compliance with every legal or technical accessibility standard.

Audience reach expansion happens when viewers can understand your video in muted environments, noisy locations, or situations where playing audio would be disruptive. Subtitles can also assist viewers who are watching in a language that is not their first.

Content discoverability may improve when a transcript or caption text gives platforms and users more information about the video. For the strongest search benefit, consider publishing a readable transcript on the page alongside the video rather than relying exclusively on a downloadable subtitle file.

Subtitles vs. Closed Captions

Although the terms are often used interchangeably, they commonly serve different purposes:

Subtitles reproduce or translate spoken dialogue for viewers who can hear the audio but need text support.
Closed captions include dialogue and meaningful non-speech information for deaf and hard-of-hearing viewers.
SDH subtitles combine dialogue with relevant sound descriptions and speaker information.

Kdenlive lets you create and edit the text needed for each approach. For projects requiring SDH subtitles, include meaningful sound cues and speaker identifications where they are needed for comprehension.

Kdenlive 101: Your Free Video Editing Companion

Kdenlive is a free, open-source nonlinear video editor licensed under the GNU GPL. It is available for Windows, macOS, Linux, and BSD.

System Requirements and Installation

Operating-system requirements can change with new releases. The current Kdenlive manual lists macOS 12 Monterey or newer for Intel-based Macs, while the latest download build may impose a newer requirement. Check the official Kdenlive download page before installation.

Speech recognition requires more storage and processing resources than basic editing. Whisper dependencies alone can require approximately 2 GB of downloads, and individual speech models may require additional space. Larger models and optional translation components require substantially more disk space, memory, and processing time.

Download Kdenlive directly from kdenlive.org and choose the installation package for your operating system.

Interface Overview

Once installed, you will find a workspace organized around:

Project Bin for organizing media files
Timeline for arranging video, audio, and subtitle content
Monitors for previewing source clips and the edited sequence
Effects Panel for applying visual and audio effects

Subtitles appear on a dedicated subtitle track in the timeline. You can open the track using the Add Subtitle or Edit Subtitle Tool icon in the Timeline Toolbar.

Method 1: Creating Subtitles Manually in Kdenlive

Manual subtitle creation gives you direct control over wording, timing, line breaks, and presentation.

Step-by-Step Manual Workflow

Open the subtitle track:

Click the Add Subtitle or Edit Subtitle Tool icon in the Timeline Toolbar.
You can also use Sequence > Subtitles to access subtitle commands.
The subtitle track will appear in the timeline.

Create individual subtitles:

Position the playhead where the dialogue begins.
Double-click the subtitle track at that position.
Enter the text directly in the subtitle item or use the Subtitle Window.
Drag the subtitle’s edges to adjust its start and end times.
Repeat the process for each section of dialogue.

You can also press Shift+S to add a subtitle at the current position.

Fine-tune timing:

Drag each subtitle to change its position.
Drag its edges to change its duration.
Zoom into the timeline when you need more precise adjustments.
Keep subtitles on screen long enough to read without allowing them to remain after the corresponding speech has ended.

Kdenlive also supports splitting subtitle text, spell-checking, ASS override tags, multiple subtitle layers, and navigation between subtitle entries.

Time Investment Reality Check

Manual subtitle creation usually takes longer than the video’s runtime because the work includes:

Listening to and transcribing the dialogue
Dividing text into readable subtitle entries
Synchronizing each entry with the audio
Reviewing spelling, punctuation, and speaker changes
Adding non-speech information when accessibility captions are required
Formatting and quality control

The total time varies significantly with audio quality, speaker count, subtitle density, and the level of accuracy required.

Method 2: Using Kdenlive’s Built-in Speech-to-Text

Kdenlive includes local automatic speech recognition through VOSK and Whisper. The feature can automatically create subtitles or generate a transcript for supported editing workflows.

Technical Setup Requirements

Before using speech recognition, open:

Settings > Configure Kdenlive > Speech to Text

Kdenlive uses a Python environment and separate language or speech models for its recognition engines.

For VOSK:

Install or configure the required Python components.
Download a compatible VOSK language model.
Add the model through Kdenlive’s model controls. You can select a downloaded model file or drag a model link into the model window where supported.

For Whisper:

Select Whisper as the speech-recognition engine.
Install the missing dependencies when prompted.
Open Manage models and install a model, or download Kdenlive’s recommended model.
Choose CPU processing or a supported CUDA-capable GPU.

Current documentation recommends the Whisper turbo model, which is approximately 1.4 GB. Whisper’s initial dependencies require an additional download of approximately 2 GB.

Optional multilingual translation through SeamlessM4T requires a much larger model of roughly 9–10 GB and can place a heavy load on the CPU, memory, and storage system.

Creating Automatic Subtitles

After configuring an engine:

Open the subtitle track from the Timeline Toolbar.
Click the Speech Recognition control to open the Automatic Subtitling dialog.
Choose the timeline zone, selected clips, or other area you want to process.
Select the speech model.
For Whisper, confirm the language or leave it on automatic detection.
Adjust options such as the maximum characters per line when needed.
Click Process.

Kdenlive creates and inserts subtitle entries automatically. Review the result for wording, punctuation, speaker changes, line breaks, and timing.

The Reality of Local AI Transcription

Local speech recognition has several advantages:

Audio can be processed without uploading it to an external transcription platform.
There is no per-minute transcription charge.
Whisper and VOSK can be used for different languages and hardware configurations.
Whisper supports language detection and translation to English.

There are also practical tradeoffs:

Dependencies and models require additional downloads.
Some models use several gigabytes of disk space.
CPU processing can be slow, particularly with larger Whisper models.
GPU acceleration requires compatible NVIDIA CUDA hardware.
Long recordings and large models can demand significant processing resources.
Every automatically generated subtitle track should be reviewed.

Kdenlive states that Whisper is generally more accurate than VOSK, while VOSK is faster on a CPU. It does not publish a universal accuracy percentage for either engine because results depend heavily on the recording and selected model.

Method 3: The Hybrid Approach External Transcription Plus Kdenlive

A hybrid workflow uses an external transcription platform to generate the initial timed text and Kdenlive to handle video editing, subtitle review, styling, and final rendering.

Why the Hybrid Workflow Can Help

An external automated transcription service can handle the initial speech-to-text processing and produce a subtitle file for Kdenlive.

Potential benefits include:

No local speech-model installation
Less demand on your computer’s CPU or GPU
Timed subtitle files ready for import
Browser-based transcript review
Multiple export formats
Simpler collaboration with reviewers

The actual amount of time saved depends on the quality and duration of the recording and the amount of editing the transcript requires.

Importing External Subtitle Files

Once you have a compatible subtitle file:

Open your sequence in Kdenlive.
Go to Sequence > Subtitles > Import Subtitle File.
Select an ASS, SRT, VTT, or SBV file.
Choose whether to create a new subtitle file or replace the current one where the dialog provides that option.
Review the imported entries in the subtitle track.
Correct the text, line breaks, timing, and styles as needed.

Kdenlive can maintain multiple subtitle files, although only one subtitle file is active at a time. The active file is the one used when the project is rendered.

Format Compatibility

Kdenlive can import:

SRT a widely supported plain-text subtitle format
ASS a format with advanced styling and positioning support
VTT a subtitle format commonly used with HTML5 video
SBV a simple timed-text format associated with online video workflows

Kdenlive’s documented export formats are:

SRT
ASS

For broad platform compatibility, converting video to SRT is usually the most practical choice.

Styling and Customizing Your Subtitles

Kdenlive uses ASS-based subtitle capabilities and offers extensive styling controls.

Available Styling Options

Depending on the selected style and ASS properties, you can control:

Font family and size
Text, outline, shadow, and background colors
Alignment and positioning
Margins
Borders and shadows
Transparency
Spacing and other typography settings

Kdenlive also supports ASS override tags for styling selected portions of a subtitle.

Creating Consistent Styles

To manage subtitle styles:

Open Sequence > Subtitles > Manage Subtitles.
Select the Styles section.
Choose the subtitle file or global style area.
Click Add Style.
Configure the style in the Style Editor.
Assign the style through the Style control in the Subtitle Window.

Styles can be created for a particular subtitle file or stored globally for reuse. You can also assign a default style to a subtitle layer, which is helpful when using different visual treatments for different speakers.

Exporting Your Subtitles from Kdenlive

After reviewing the subtitle track, you can export it as a separate file or include it when rendering the video.

Separate Subtitle Files

Go to:

Sequence > Subtitles > Export Subtitle File

Kdenlive can export the active subtitle file as:

SRT
ASS

SRT is widely accepted by video-hosting and publishing platforms. ASS is useful when you need to preserve more advanced styles.

Burned-in Subtitles

Burned-in subtitles are permanently rendered into the video image and cannot be turned off.

To burn subtitles into the video:

Confirm that the correct subtitle file is active.
Open the Render dialog.
Open the additional render options if they are hidden.
Leave Embed subtitles instead of burning them in unchecked.
Render the video.

Kdenlive will render the active subtitles into the video image.

Embedded or Soft Subtitles

Kdenlive can also create a subtitle stream inside a compatible media container.

To embed subtitles:

Choose a render format and container that support subtitle streams, such as MKV.
Enable Embed subtitles instead of burning them in.
Render the file.

Embedded subtitles can be turned on or off in a compatible player. Compatibility depends on the selected container, codec, playback application, and distribution platform.

Why Sonix Makes Kdenlive Subtitles Easier

Kdenlive can handle the editing, styling, and rendering stages, while Sonix can create the initial transcript and timed subtitle file.

Processing designed for recorded content: Sonix says it typically processes one hour of audio or video in approximately five minutes. Processing time can vary with file size and service load.

Accuracy that can reduce cleanup: Sonix advertises up to 99% transcription accuracy for clear recordings. Results vary with background noise, overlapping speech, speaker clarity, accents, microphone quality, and specialized vocabulary. Its Custom Dictionary feature allows users to add names and industry-specific terms.

Multi-language support: Sonix supports transcription in 54+ languages and automated translation into 55+ languages.

Security controls for professional workflows: Sonix states that it is SOC 2 Type II certified and uses TLS for data in transit and AES-256 server-side encryption for data at rest. More information is available on the Sonix security page.

Format flexibility: Sonix can export subtitle files including SRT and VTT. For Kdenlive, SRT is the simplest choice when you want a format that the editor can both import and export.

The workflow is:

Upload your recording to Sonix.
Review the transcript in the browser editor.
Correct names, terminology, punctuation, and speaker labels.
Export the subtitles as SRT.
Import the SRT file through Sequence > Subtitles > Import Subtitle File in Kdenlive.
Make final timing and styling adjustments.
Export the subtitle file or render the video with burned-in or embedded subtitles.

Final Verdict: Choosing the Right Subtitle Workflow for Kdenlive

The best method depends on your privacy requirements, available hardware, technical comfort, project volume, and turnaround needs.

Choose manual subtitle creation when you:

Need direct control over every word, timing decision, and line break
Have a short project or an existing transcript
Need detailed accessibility captions with carefully authored sound descriptions
Cannot use automatic speech recognition for the material

Choose Kdenlive’s built-in speech-to-text when you:

Prefer to process recordings locally
Are comfortable installing dependencies and speech models
Have suitable storage and processing resources
Can review and correct automatically generated subtitles
Want to avoid uploading the recording to an external platform

Choose the Sonix + Kdenlive hybrid workflow when you:

Want to avoid configuring local speech-recognition models
Need timed subtitle files for several projects
Want a browser-based transcript-review workflow
Require multi-language translation
Value documented security controls for sensitive content
Need SRT files that can be imported directly into Kdenlive

For many professional workflows, the hybrid method offers a practical division of labor: Sonix creates the transcript and initial timing, while Kdenlive provides detailed subtitle editing, styling, video integration, and rendering.

Frequently Asked Questions

Is Kdenlive truly free, and what operating systems does it support?

Yes. Kdenlive is free, open-source software licensed under the GNU GPL. It is available for Windows, macOS, Linux, and BSD. Operating-system requirements vary by release, so check Kdenlive’s official download page before installing the latest version.

What’s the difference between hardcoded and soft subtitles when exporting?

Hardcoded or burned-in subtitles become part of the video image and cannot be disabled. Soft subtitles remain as a separate file or subtitle stream that a viewer can turn on or off in a compatible player. Kdenlive can burn the active subtitle file into a rendered video or embed it in a compatible container by selecting “Embed subtitles instead of burning them in” in the Render dialog.

Can I use external transcription services with Kdenlive?

Yes. Kdenlive can import ASS, SRT, VTT, and SBV subtitle files. Generate and review the transcript in the external service, export a compatible file such as SRT, and import it through Sequence > Subtitles > Import Subtitle File.

Why does Kdenlive’s speech-to-text feature require additional setup?

Kdenlive performs speech recognition through separate engines such as VOSK and Whisper. These engines require Python components, dependencies, and downloadable speech models that are not included in the basic editor installation. Kdenlive provides configuration and model-management controls under Settings > Configure Kdenlive > Speech to Text.

How accurate are Kdenlive’s built-in transcription models?

Kdenlive does not publish a universal accuracy percentage for VOSK or Whisper. Its documentation describes Whisper as generally more accurate than VOSK and VOSK as faster on a CPU. Actual results depend on the language, selected model, background noise, recording quality, number of speakers, and whether speakers overlap.

Get accurate transcription in minutes

Start transcribing smarter. Try Sonix free or explore our pricing to find the right plan for you.

Try Sonix Free See Pricing