In 2026, two types of captions dominate video content. The first is word-by-word animated captions — the karaoke-style text where each word highlights as it's spoken, the format that drives millions of views on Reels, Shorts, and TikTok. The second is accurate multi-language subtitles for global reach: the same video captioned in Spanish, Hindi, Korean, and nine other languages without a translation agency.
DaVinci Resolve's built-in subtitle tool does neither of these automatically. It lets you type subtitle text yourself — every line, by hand, no AI transcription. It's a genuinely useful tool if you already have a transcript and want to sync it manually. But for most creators in 2026, manual subtitling is a bottleneck.
DavinciClaude fills both gaps with two dedicated features: Smart Captions for word-by-word animated captions, and Smart Subtitles for 99-language subtitle generation. Both work natively inside DaVinci Resolve. Both are free. This guide covers how each one works, when to use which, and how they compare to DaVinci's native tools and web-based alternatives like Captions.ai and SubMagic.
What DaVinci Resolve's native subtitle tool actually does
Before covering DavinciClaude's tools, it's worth being clear about what DaVinci Resolve ships with — because the honest answer is that it's more limited than most people expect.
In the Edit page, you can add a subtitle track via Timeline > Add Subtitle Track. This gives you a dedicated subtitle lane where you can create subtitle clips, type text, and time them to your audio. You can adjust font, size, position, and export the subtitle track as a .srt file or as embedded metadata on your export.
What it cannot do: transcribe your audio. There is no AI transcription built into DaVinci Resolve's subtitle tool. You're typing every word yourself, then manually adjusting timing. For a 10-minute interview, that's 45–90 minutes of work. For a 1-hour video, plan on 3–4 hours.
DaVinci Resolve also has no built-in word-by-word caption mode — the animated karaoke style where each individual word appears in sequence. That format isn't supported natively at all.
Part A: Word-by-word animated captions with Smart Captions
What word-by-word captions are (and why they work)
Word-by-word captions are exactly what they sound like: instead of displaying a full line of subtitle text, each word appears and highlights as the speaker says it. The previous word stays visible for a moment, then the new word pops up and lights up. This creates a karaoke-style reading experience that keeps viewers engaged even on muted autoplay.
Every viral talking-head video on Instagram Reels uses this format. Every podcast clip that racks up 500k views is doing this. It's not a trend that's going away — viewer attention data consistently shows higher watch-through rates on word-by-word caption content compared to static subtitles, because the text movement keeps the eye active.
Creating this format manually in DaVinci Resolve without a plugin means creating a separate title clip for every single word, timing each one to the audio, and then formatting them consistently. For a 60-second clip with normal conversational pacing (~150 words per minute), that's 150 individual title clips to create and time. Nobody does this by hand.
How Smart Captions works
Smart Captions automates this entirely using a two-step AI process. First, Whisper transcribes the audio to produce a precise word-level transcript — including timestamps for the start and end of every individual word. Then Claude refines the transcript, catching the errors that pure transcription models miss: proper nouns spelled incorrectly, technical terminology, mid-sentence corrections where the speaker stumbled. The combined accuracy rate is 99.5%.
The output is placed on your DaVinci Resolve timeline as individual native title clips — one per word. Each clip is a real, editable Resolve title. You can click any word to change its text, restyle the font, move it, or delete it. Nothing is baked. If Smart Captions mistranscribed a name or technical term, you click the word clip and fix it in five seconds.
Step-by-step: generating word-by-word captions in DaVinci Resolve
- Open Smart Captions. In DaVinci Resolve, go to Workspace > DavinciClaude > Smart Captions.
- Select a style. Choose from preset caption styles — bold white text on a dark pill background, minimal text only, outlined, or custom. The style controls the visual appearance of every caption clip that gets placed on your timeline.
- Click Generate. DavinciClaude transcribes the audio and places word clips on your timeline. For a 10-minute clip, this takes roughly 60–90 seconds.
- Captions appear on your timeline. Each word is a separate title clip in a new track above your video, timed precisely to when the word is spoken.
- Click any word to edit it. If a name or technical term is wrong, click that specific clip and retype it. Changes don't affect any other word. You're editing native DaVinci Resolve title clips — no special workflow required.
You can also use Copilot to run Smart Captions with natural language: open the Copilot panel and type "add word-by-word captions to the selection" and it handles the rest.
Why this beats web-based tools
Tools like Captions.ai and SubMagic generate word-by-word captions in a browser — you upload your video, they transcribe it, and you download a new video file with captions burned in. The problem is the output is a baked render: you can't change the font, move the captions, or fix individual words without re-running the whole process. And because the captions are burned into a new video file, if you later change your edit you have to re-caption everything from scratch.
Smart Captions outputs editable title clips that live on your Resolve timeline. Your edit is still the source of truth. The captions travel with your project, not with an exported file. When you make a cut or trim a clip, the caption clips stay in sync with your edit.
Part B: 99-language subtitle generation with Smart Subtitles
What Smart Subtitles does
Smart Subtitles handles a different use case: generating accurate subtitle tracks in multiple languages for international distribution. Instead of word-by-word animation, this is traditional subtitle formatting — line-wrapped text that appears during the spoken phrase, the kind you see on YouTube, Netflix, or any YouTube video with CC enabled.
The workflow starts with Whisper transcribing your English audio to produce a precise transcript. That transcript is then translated into any of 99 supported languages, and the result is formatted into properly timed subtitle segments. The output is a native DaVinci Resolve subtitle track that you can keep on your timeline or export as a .srt file.
If you create content with an international audience — or want to reach one — this is the feature that makes it tractable. A single recording in English can become subtitled content in Spanish, French, Portuguese, Hindi, Korean, Japanese, and German in about the time it takes to make coffee.
Step-by-step: generating multi-language subtitles in DaVinci Resolve
- Open Smart Subtitles. Go to Workspace > DavinciClaude > Smart Subtitles.
- Select your target language(s). Choose one or multiple languages from the 99 available. Common selections: Spanish, French, Portuguese, Hindi, Korean, Japanese, German, Arabic.
- Click Generate. DavinciClaude transcribes the audio and translates to each selected language simultaneously. Processing time scales with the number of languages and video length.
- Subtitle tracks appear on your timeline. Each language gets its own subtitle track in Resolve's subtitle lane. You can toggle visibility per language.
- Export or burn in. Export the subtitle track as a .srt file for YouTube or Vimeo upload (where the viewer can toggle subtitles on/off), or burn it into the video for platforms that don't support external subtitle files.
As with Smart Captions, you can also trigger Smart Subtitles via Copilot: "add French subtitles" or "generate Spanish and Portuguese subtitles" both work as natural-language instructions.
Speaker-aware captions with Diarization
If your video has multiple speakers — a podcast, interview, or panel discussion — Diarization adds speaker identification to your captions. Each speaker gets a distinct color, so viewers can track who's saying what even without video switching. This works alongside both Smart Captions and Smart Subtitles.
Honest comparison: DaVinci native, DavinciClaude, Captions.ai, SubMagic
Here's how the options stack up across the factors that matter:
| Feature | DaVinci Native | Smart Captions | Smart Subtitles | Captions.ai | SubMagic |
|---|---|---|---|---|---|
| AI transcription | ❌ | ✅ | ✅ | ✅ | ✅ |
| Word-by-word animated | ❌ | ✅ | ❌ | ✅ | ✅ |
| Native in Resolve | ✅ | ✅ | ✅ | ❌ | ❌ |
| Languages supported | 1 (manual) | 1 | 99 | ~12 | ~15 |
| Translation built-in | ❌ | ❌ | ✅ | Limited | Limited |
| Editable as title clips | ✅ (SRT only) | ✅ (per word) | ✅ (SRT / track) | ❌ (baked render) | ❌ (baked render) |
| Stays in your NLE | ✅ | ✅ | ✅ | ❌ | ❌ |
| Price | Free | Free | Free | $18/mo | $20/mo |
The web-based tools (Captions.ai and SubMagic) have a fundamental workflow problem: they require you to leave DaVinci Resolve, upload your video, download a captioned version, and bring it back. If your edit changes — a common scenario — you repeat the entire process. The output is also a baked render, not editable title clips, which means you're stuck with whatever font and position they chose.
The tradeoffs are real: Captions.ai has a polished UI for caption styling, and SubMagic is specifically optimized for short-form social formats. But both cost money every month for something that DavinciClaude does for free, inside Resolve, with fully editable output.
A note on accuracy: 99.5% and what it means in practice
Smart Captions achieves 99.5% word-level accuracy on clean audio. That number comes from the combination of Whisper's transcription and Claude's refinement pass — Whisper produces a strong initial transcript, and Claude catches the errors that pure audio models typically miss: proper nouns, technical terminology, brand names, and words that sound similar but differ by context.
In practice, 99.5% on a 10-minute video (roughly 1,500 words) means about 7–8 words will be wrong. Most of those will be niche proper nouns or highly technical terms. The fix is always the same: click the title clip for that word and retype it. Because Smart Captions outputs individual editable clips, you correct one word without touching anything else.
For non-English audio: Smart Subtitles handles transcription in languages other than English. Accuracy varies by language — high-resource languages (Spanish, French, German, Japanese) perform similarly to English. Lower-resource languages may have slightly lower accuracy on technical content.
Frequently asked questions
How do I edit captions after they're generated?
Click directly on the title clip for the word you want to change in the DaVinci Resolve timeline. This opens the standard Resolve title editor where you can retype the word. Changes apply only to that clip and don't affect timing or any surrounding words. You edit Smart Captions output exactly the same way you'd edit any other DaVinci Resolve title.
Can I change the font and style of the captions?
Yes. Each word is a native DaVinci Resolve title clip, so you can apply any font, color, size, drop shadow, or background styling that Resolve supports. You can also select all caption clips at once and change properties in bulk. If you want a consistent style applied at generation time, choose a preset in the Smart Captions panel before clicking Generate.
Does Smart Captions work with non-English audio?
Smart Captions transcribes in the language of the audio. Whisper supports a wide range of languages natively, so English, Spanish, French, German, Japanese, Portuguese, and most widely spoken languages produce good results. If you want subtitles in a different language than your audio — for example, Spanish subtitles on English audio — use Smart Subtitles with translation enabled.
How many languages does Smart Subtitles support?
99 languages for both transcription and translation. The full list includes all major world languages and a significant number of regional ones. You can generate multiple language versions from a single English recording in one pass — select all the languages you need and DavinciClaude generates all of them simultaneously.
Is DavinciClaude free?
Smart Captions and Smart Subtitles are both free features of DavinciClaude. You can generate word-by-word captions and 99-language subtitles at no cost. Download DavinciClaude to get started — the install takes under 60 seconds and works with DaVinci Resolve 18 and later.
DaVinci Resolve is an excellent NLE that ships with a genuinely limited subtitle tool. In 2026, that means every DaVinci Resolve user who wants word-by-word animated captions or multilingual subtitle tracks has to either type everything manually or leave Resolve to use a web-based tool and bring a baked file back into their project.
DavinciClaude gives you both features natively — word-by-word animated captions as individual editable title clips, and 99-language subtitle generation as native Resolve tracks. Both free. No upload, no download, no rebake.
→ Download DavinciClaude and add your first AI-generated captions in under two minutes.



