--- title: Transcribe and subtitle display: '*Transcribe* and subtitle' nav_title: Transcribe and subtitle summary: - "Get a transcript, an SRT file or burned-in subtitles from any recording, and fix them before they ship." - "Choose the output first; prefer an SRT over burned-in text when the platform accepts files." - "Transcription starts in Create > Text; a short transcript is one of the cheapest runs and checks any generated speech." description: Turn speech into text and SRT subtitles, burn styled captions into a video, translate them, and fix the transcript before it ships. section: audio order: 70 type: guide level: Start roles: [creator, artist, art-director, marketer, team-lead, admin, developer] time: { read: 8, do: 10 } needs: ["Any plan with credits", "An audio file or a video with speech"] last_verified: 2026-11-11 models_checked: ["Gemini 3.5 Transcribe", "Speech to Text", "Auto Subtitles", "Scenario Caption Studio"] related: [narration-and-speech] pdf: true thumb: audio/transcribe-and-subtitle/thumb.webp video: tutorials/transcribe-and-subtitle.mp4 chapters: - { at: 0, title: "Text and an SRT", image: "/media/images/audio/transcribe-and-subtitle/ch-01-text-and-an-srt.webp" } - { at: 16, title: "Open the model, add language hints", image: "/media/images/audio/transcribe-and-subtitle/ch-02-open-the-model-add-language-hints.webp" } - { at: 33.5, title: "Choose the settings and generate", image: "/media/images/audio/transcribe-and-subtitle/ch-03-choose-the-settings-and-generate.webp" } - { at: 48.5, title: "Fix the SRT by hand", image: "/media/images/audio/transcribe-and-subtitle/ch-04-fix-the-srt-by-hand.webp" } - { at: 77, title: "Burn subtitles into the video", image: "/media/images/audio/transcribe-and-subtitle/ch-05-burn-subtitles-into-the-video.webp" } - { at: 100.5, title: "Prefer a file where you can", image: "/media/images/audio/transcribe-and-subtitle/ch-06-prefer-a-file-where-you-can.webp" } - { at: 108.5, title: "Recap", image: "/media/images/audio/transcribe-and-subtitle/ch-07-recap.webp" } steps: - title: "Open the transcription model" text: "Create › Text › Gemini 3.5 Transcribe. Under Audio, drag and drop the recording, click Select from Library, or Upload." - title: "Help it with names and language" text: "In Language hints, type a code such as en-US and click Add; in Custom vocabulary, add names and invented words one at a time, here Keeper. Keep the list to distinctive terms." - title: "Choose the settings and generate" text: "Under Settings, turn on Speaker diarization for dialogue and Word timestamps for subtitles; either one lowers the longest file from 60 minutes to 30. Generate gives a text file and an SRT." - title: "Fix the SRT by hand" text: "Open the SRT in any text editor, fix words in place and split long subtitles at a breath, each on screen for about seven seconds at most, with its own number and time range. Write numbers one way throughout." - title: "Burn subtitles into the video" text: "Pick Auto Subtitles or Scenario Caption Studio and add the video under Input Video. In Caption Studio: Style Preset Modern Chip, Translation › Target Language Auto, and Subtitles (SRT) for a fixed SRT; then watch the result on the target device." - title: "Prefer a file where you can" text: "If the platform takes an SRT, deliver the SRT: you can fix it any time. Burned-in subtitles can't be turned off or corrected without generating the video again." agent_steps: - title: "Connect the Scenario MCP" text: "Connect your agent and sign in once at [mcp.scenario.com](https://mcp.scenario.com). Then tell it which Scenario team and project to use, so every file lands there." - title: "Give the agent the recording" text: "The audio or video to transcribe: upload it with upload_asset or give its asset ID. Add the names and invented words it should spell your way." - title: "Brief the agent: the transcript and the SRT" prompt: |- Transcribe my recording with Gemini 3.5 Transcribe (model_google-gemini-3-5-transcribe), reading model_schema_get first for the parameter names. Language hint en-US, custom vocabulary: Keeper. Word timestamps on; speaker diarization off (on if it's a dialogue). I want the text and the SRT. Quote the cost with a dry run first and wait for my go. - title: "Fix the SRT" prompt: |- Wait for the job with jobs_wait and download the SRT with asset_download. Show me the transcript against my script, if I gave you one, and list every word that differs. Then split every subtitle longer than about seven seconds at a comma or a breath, one or two short lines each, keep the numbering in order, and write numbers one way throughout. Show me the fixed SRT before anything else. - title: "Burn in, only if the platform needs it" text: "Deliver the SRT file when the platform takes one. Burn in only when it can't, and watch the result on the target device." prompt: |- Burn the subtitles into my video with Scenario Caption Studio (model_scenario-caption-studio), reading model_schema_get first. Style preset Modern Chip, target language Auto. If the schema takes an SRT, upload my fixed SRT with upload_asset and use it for the text and timing. Quote the cost with a dry run first and wait for my go. - title: "Check and report" text: "Accept when every word matches, names are spelled your way, numbers are written one way and each subtitle is short enough to read. Reject a wrong or missing word, swapped speaker labels or burned-in text over an important part of the picture." prompt: |- Create a collection with the recording, the transcript, the fixed SRT and the subtitled video, and report the models, the settings and every asset ID. updated: 2026-11-11 canonical: https://learn.scenario.com/audio/transcribe-and-subtitle --- Last updated: 2026-11-11 # Transcribe and subtitle > Turn speech into text and SRT subtitles, burn styled captions into a video, translate them, and fix the transcript before it ships. In short: - Get a transcript, an SRT file or burned-in subtitles from any recording, and fix them before they ship. - Choose the output first; prefer an SRT over burned-in text when the platform accepts files. - Transcription starts in Create > Text; a short transcript is one of the cheapest runs and checks any generated speech. This lesson gets the words out of a recording: a transcript, a subtitle file, or subtitles burned into the video. You'll transcribe the lighthouse narration with Gemini 3.5 Transcribe, fix the SRT file by hand, and burn subtitles into a video with Scenario Caption Studio. It's also the fastest way to check any speech you generate. :::note As of November 2026 The models, settings and screens in this article are the ones Scenario offered in November 2026. New models and updates arrive often, so check the app for newer options before you choose. ::: :::steps Quick version 1. Open **Create** › **Text** and pick **Gemini 3.5 Transcribe**. Add your audio or video under **Audio**. 2. Add a **Language hints** entry if you know the language, and **Custom vocabulary** for names and invented words. 3. Turn on **Speaker diarization** for dialogue and **Word timestamps** for subtitles, then click **Generate**: you get a text file and an SRT. 4. Fix the SRT in any text editor: words in place, long subtitles split, numbers written one way. 5. To burn subtitles in, use **Auto Subtitles** or **Scenario Caption Studio**, then watch the result on the target device. ::: ## Before you start Pick the output first, then the model: | You need | Model | What you get | |---|---|---| | A transcript and an SRT, with speaker labels or word timings | Gemini 3.5 Transcribe | A text asset and an SRT asset. Up to 60 minutes of audio, or 30 minutes with speaker labels or word timestamps on | | A quick transcript or SRT, or an English translation | Speech to Text | Text or SRT; detects the language and can translate to English | | Subtitles burned into a video, with simple styling | Auto Subtitles | The video with subtitles: font, colors, border, line length, language | | Styled captions, translated, burned in or as a file | Scenario Caption Studio | Captions in one of 7 style presets or your own style, translation into 18 languages, burned in or as an SRT | Transcription lives under **Create** › **Text** ("Transcribe audio, summarize images, or run text-to-text models"), not under Audio. :::flow From recording to subtitles Transcribe | Gemini 3.5 Transcribe, text and SRT | text Fix | names, numbers, long lines | text Deliver | SRT file, or burn in | video Check | on the target device | ::: ## Step 1: Open the transcription model 1. Open **Create** › **Text**. 2. Pick **Gemini 3.5 Transcribe** in the model menu. 3. Under **Audio**, add your recording: drag and drop it, click **Select from Library**, or **Upload**. Here, the ElevenLabs v4 narration from [Narration and speech](https://learn.scenario.com/audio/narration-and-speech.md). ## Step 2: Help it with names and language - **Language hints:** type a language code and click **Add**, here `en-US`. Leave it empty and the model detects the language among more than 85, even when a speaker switches mid-sentence. A hint helps most on short clips, accents and close dialects. - **Custom vocabulary:** add names, product terms and invented words, one at a time with **Add**, so they're spelled your way. Here, `Keeper`. Keep the list to distinctive terms; everyday words weaken it. ## Step 3: Choose the settings and generate Under **Settings**: - **Speaker diarization** labels who speaks (speaker_0, speaker_1 and so on). Turn it on for interviews and dialogue. - **Word timestamps** splits the transcript into timed, numbered segments. Turn it on if you'll fine-tune subtitles. Here it's on. Either setting lowers the longest file the model accepts from 60 minutes to 30. Click **Generate**. You get two files in your session: the text, and an SRT subtitle file. ![Gemini 3.5 Transcribe with the narration under Audio, en-US in Language hints, Keeper in Custom vocabulary, Word timestamps on, and the text and SRT results in the session](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/af7aec401da8d033c9d43d2fb795589081193c6fcb27e90bae48c18661c20606.webp) ## Step 4: Fix the SRT by hand An SRT is plain text: a number, a time range, and the words, then a blank line, repeated. Open it in any text editor. The narration came back as one subtitle: ```text 1 00:00:00,300 --> 00:00:12,500 Every evening at six, the keeper climbs the 112 steps of the lighthouse. The lamp is old, the stairs are older, and the cat, as usual, is already at the top. ``` Every word matches the script, but the subtitle is far too long to read. To fix it: - **Fix words in place** and keep the time lines. - **Split long subtitles** at a comma or a breath: one or two short lines each, on screen for no more than about seven seconds. Give each part its own number and time range. - **Keep the numbering in order** after a split. The same narration, split at a breath: ```text 1 00:00:00,300 --> 00:00:05,900 Every evening at six, the keeper climbs the 112 steps of the lighthouse. 2 00:00:06,300 --> 00:00:12,500 The lamp is old, the stairs are older, and the cat, as usual, is already at the top. ``` **Decide how numbers are written.** The narration said "a hundred and twelve"; the transcript wrote the digits. Either is fine, as long as it's the same everywhere. ## Step 5: Burn subtitles into the video When a platform can't take a subtitle file, burn the subtitles into the video with **Auto Subtitles** or **Scenario Caption Studio**. Find them by name in the model menu, and add the video under **Input Video**. In Scenario Caption Studio: - **Style Preset:** the look of the captions. The film uses **Modern Chip**, a rounded background behind the text. The others include TikTok Bouncy, Karaoke Fill, Karaoke Underline and Word Pop. - **Translation** › **Target Language:** **Auto** keeps the spoken language; pick another to translate. - **Text Segments:** **Add Text Segments** puts your own timed lines on screen instead of transcribing, for a title or a hook. - **Subtitles (SRT):** use a fixed SRT for the text and timing, or get the SRT file out with the video. - Segment limits: shorter segments, with fewer words on screen at once, are easier to read on a phone. ![Scenario Caption Studio settings: Subtitles (SRT), Text Segments, Translation with Target Language Auto, and Style with Style Preset Modern Chip](https://ufgc0c9mhl4ueq76.public.blob.vercel-storage.com/academy/757adb53123668f9a70d3b4a9af7fea6fa56b0c86840bae53d01a5753b0ffc7e.webp) Auto Subtitles is the simpler choice: a fixed style you adjust with font, size, colors and border, the original language or a translation to English, and limits on how long a line stays on screen. Generate, then **watch the result on the device it's made for**: a phone for vertical video, a TV or laptop for a trailer. ## Step 6: Prefer a file where you can If the platform takes an SRT, deliver the SRT. You can fix it any time, and players and platforms load it. Burned-in subtitles can't be turned off or corrected without generating the video again. ## Use transcription as a quality check Transcribe every generated speech take and every instrumental: - **Speech:** compare the transcript with the script word by word. A changed word, such as "night house" for "lighthouse" after a voice change, is easy to read and hard to hear. - **Music:** an instrumental should return no speech. Any text means the model sang or spoke something. :::check **Accept when:** every word matches what was said, names and terms are spelled the way your project writes them, speakers are labeled correctly, numbers are written one way throughout, and each subtitle is short enough to read in the time it's on screen. **Reject when:** a word is wrong or missing, a name is spelled differently between subtitles, speaker labels swap, a subtitle stays on screen for more than about seven seconds, or burned-in text covers an important part of the picture. ::: ## Common problems | What you see | What to do | |---|---| | Names come out wrong | Add them to **Custom vocabulary** and run again | | The wrong language is detected | Add a **Language hints** entry; short clips are the most likely to be misdetected | | Speaker labels are mixed up | Overlapping speech confuses labeling; clean the audio with ElevenLabs Voice Isolator first, or fix labels by hand | | A long file is rejected | Split it, or turn off **Speaker diarization** and **Word timestamps**, which halve the limit | | One subtitle runs for the whole clip | Turn on **Word timestamps**, or split the SRT by hand at commas and breaths | | Subtitles drift after an edit | The video was trimmed after transcription; transcribe the final cut | | A typo is burned into the video | Fix the SRT, then generate again from it with **Subtitles (SRT)** | | The job failed or is stuck | See [Failed or stuck jobs](https://learn.scenario.com/browse#help-center) | ## What's next That's the course: music, voice and sound on Scenario, and every take checked before it ships. To go back over model choice and prompts, see [Audio on Scenario](https://learn.scenario.com/audio/audio-on-scenario.md); to automate transcripts for a series, put Gemini 3.5 Transcribe in a [workflow](https://learn.scenario.com/workflows/what-are-workflows.md). Source: https://learn.scenario.com/audio/transcribe-and-subtitle