Twitch Video Transcript
Pricing
from $0.378 / transcript
Twitch Video Transcript
Twitch Video Transcript captures dialogue from one public Twitch Clip for highlight review, creator research, and quote extraction. Receive detected-language text, ordered timestamps, clip metadata, and optional translation into 133 languages. The lowest transcript tier is $0.378.
Pricing
from $0.378 / transcript
Rating
0.0
(0)
Developer
TrueFetch
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
2 hours ago
Last modified
Categories
Share
Affiliate disclosure: Apify links on this page may include referral parameters. If you sign up through one of these links, TrueFetch may earn a commission from Apify at no extra cost to you. Pricing, features, and Actor access are unaffected.
Twitch Video Transcript converts one public Twitch Clip into detected-language text, timestamped segments, clip metadata, and an optional translated transcript.
- The required input is one public Twitch Clip URL; channel pages, live streams, VOD pages, Clip Manager pages, uploads, and cross-platform links are outside this Actor's scope.
- One completed Clip writes one 22-field item that joins streamer speech, phrase timing, channel context, game/category data, and exposed engagement values.
- All 133 translation targets are optional; the extra event is emitted only for a complete time-aligned target transcript.
- The precheck resolves the Clip slug and confirms Twitch before media download and speech recognition begin.
Run a one-result test · View API Use the prefilled Twitch Clip for a minimal FREE-tier trial. Fixed events are $0.42 for transcription and $0.008 to start the default 8 GB Actor; runtime use varies. Finished Spanish output adds $0.12.
What does Twitch Video Transcript do?
Twitch Video Transcript turns speech from one public Twitch Clip into detected-language text and ordered HH:MM:SS.mmm segments. When exposed, clip metadata can include title, description, creator, duration, thumbnail, publishing time, and view count.
translate creates a companion language version for the Clip. The detected streamer speech remains under transcript, and target phrases retain matching intervals only when every segment completes.
It does not search channels, monitor a live broadcast, process full VODs, discover Clips, enumerate chat, label speakers, download video, or create subtitle files. The job is one known public Clip slug.
How do I run Twitch Video Transcript?
- Open Twitch Video Transcript on Apify.
- Copy the share URL of one public Twitch Clip and paste it into Video URL.
- Pick Translate only when a target-language copy is needed; otherwise retain source speech alone.
- Click Start and wait for media preparation and speech recognition to finish.
- Review the Dataset item for streamer names, game jargon, nullable fields, and timing before export.
Start with a Twitch Clip containing clear dialogue. The current input schema pre-fills this public URL:
{"video_url": "https://clips.twitch.tv/FaintLightGullWholeWheat","translate": "spanish"}
That Clip remains the entire run scope and yields at most one item. Twitch or the broadcaster can remove it later.
What data does Twitch Video Transcript return?
The Actor writes one successful result to the default Dataset. Its 22-field contract is nullable where Twitch omits a value.
| Group | Dataset fields | Meaning |
|---|---|---|
| Processing | processor, processed_at | Actor URL and UTC processing time |
| Twitch Clip | platform, title, description, thumbnail, published_at | Source identity, text, image, and publishing time |
| Creator | author, author_id, author_url | Creator details exposed for the media |
| Media | duration, audio_title, audio_artist | Source-reported length and audio attribution |
| Engagement | view_count, like_count, shares_count, dislike_count, comment_count | Clip counters the public response exposes; unavailable measures stay nullable |
| Labels | categories, tags | Source-provided classifications when available |
| Speech | transcript, translation | Original timestamped transcript and optional translated version |
This abbreviated response uses the same Twitch-Clip-and-Spanish scenario. Values omitted by Twitch remain null, zero, or empty; the Actor does not invent them.
{"processor": "https://apify.com/truefetch/twitch-video-transcript?fpr=aiagentapi","processed_at": "2026-07-21T13:00:03+00:00","platform": "TwitchClips","title": "EA Play 2016 Live from the Novo Theatre","description": null,"author": "stereotype_","author_id": "EA","duration": 32,"view_count": 30285,"transcript": {"language": "English","text": "SMG, nice close-quarters weapon. That's right, using it to full effect.","segments": [{"start": "00:00:00.000","end": "00:00:09.810","text": "SMG, nice close-quarters weapon. That's right, using it to full effect."}]},"translation": {"language": "Spanish","text": "SMG, una buena arma de corto alcance. Así es, usándola con plena eficacia.","segments": [{"start": "00:00:00.000","end": "00:00:09.810","text": "SMG, una buena arma de corto alcance. Así es, usándola con plena eficacia."}]}}
The contract excludes the Clip file, channel or chat history, private account data, summaries, confidence values, speaker labels, SRT/VTT output, and word-level cues.
What inputs can I configure?
| Input | Type | Required | Behavior |
|---|---|---|---|
video_url | string | Yes | One public Twitch Clip share URL. |
translate | string enum | No | One of 133 target-language values. Empty means no translation. |
video_url must point to one published clips.twitch.tv/{slug} page. The form has no login cookie, OAuth token, channel name, Clip search, VOD selector, upload, time filter, or URL batch.
Language values use lowercase names such as spanish, french, japanese, arabic, chinese (simplified), and chinese (traditional). Choosing one adds output and never overwrites the recognized Clip.
What platforms and markets does Twitch Video Transcript cover?
This Twitch-only Actor is tested with a public clips.twitch.tv/{slug} URL. Twitch documents that publishing a Clip provides a URL that can be copied and shared. The slug identifies one Clip rather than a channel, live stream, or full VOD.
Deleted, expired, removed, restricted, or still-unpublished Clips, channel pages, live streams, VOD pages, Clip Manager pages, and URLs without downloadable media are not promised.
Speech language is detected from audio. Review generated text and translation for streamer names, game terminology, accents, slang, and code-switching.
Process only Clips you are authorized to use and review the current Twitch Terms of Service for the intended workflow.
Why use Twitch Video Transcript?
| Verified capability | Practical benefit |
|---|---|
| Twitch-specific platform validation | Reject unrelated platform links before transcription begins. |
| Streamer speech captured as text and phrase intervals | Find a reaction, callout, explanation, or moment without replaying the entire Clip. |
| Channel and Clip context beside the words | Preserve available broadcaster, game, title, thumbnail, duration, date, and views. |
| Target-language phrases mapped to original timing | Review a highlight across languages while retaining the detected Twitch dialogue. |
| One highlight-shaped result | Feed a selected Clip into moderation review, esports research, editorial reuse, or an AI step. |
Twitch controls Clip access and visible metadata. Recognition may confuse handles, rapid game callouts, overlapping voices, sound effects, or domain jargon, so verify consequential wording against the Clip.
Who is Twitch Video Transcript for?
The Actor serves developers, streamers, esports teams, media analysts, researchers, and AI builders with a specific public Clip. Typical uses include searching commentary, reviewing highlights, drafting translations, and supplying timed text downstream.
Use another product to discover highlights, monitor streams, process VODs, collect chat, access private data, save media, read screen overlays, create legal transcripts, or work with another platform.
How can I use Twitch Video Transcript through the API or MCP?
Call truefetch/twitch-video-transcript. A short Clip may fit the synchronous Dataset route, while reliable automation should run asynchronously and poll. Apify notes that the HTTP wait can expire before processing ends.
curl -L "https://api.apify.com/v2/actors/truefetch~twitch-video-transcript/run-sync-get-dataset-items" \-H "Authorization: Bearer YOUR_APIFY_TOKEN" \-H "Content-Type: application/json" \-d '{"video_url": "https://clips.twitch.tv/FaintLightGullWholeWheat","translate": "spanish"}'
Apify's hosted MCP server can expose only this Actor to a compatible client:
{"mcpServers": {"apify": {"url": "https://mcp.apify.com?tools=truefetch/twitch-video-transcript","headers": {"Authorization": "Bearer YOUR_APIFY_TOKEN"}}}}
Once MCP is configured, send the Clip and spanish to truefetch/twitch-video-transcript, then consume the single Dataset item. See the Apify MCP guide for setup and the Actor API page for clients.
How much does Twitch Video Transcript cost?
Start, runtime usage, completed transcript, and optional completed translation are separate events. Compare the local values below with the live pricing page before scheduling many Clips.
| Billed event | FREE | BRONZE | SILVER | GOLD | PLATINUM | DIAMOND | Billing unit |
|---|---|---|---|---|---|---|---|
| Actor start | $0.001 | $0.001 | $0.001 | $0.001 | $0.001 | $0.001 | One event per allocated GB, minimum one |
| Actor usage | $0.00001 | $0.00001 | $0.00001 | $0.00001 | $0.00001 | $0.00001 | Runtime, proxy, and storage usage unit |
| Transcript | $0.42000 | $0.40600 | $0.39200 | $0.37800 | $0.37800 | $0.37800 | One successful timestamped transcript |
| Translation | $0.12000 | $0.11600 | $0.11200 | $0.10800 | $0.10800 | $0.10800 | One complete translated transcript when requested |
At FREE, eight $0.001 start units represent the default 8 GB allocation. One $0.42000 transcript makes $0.428 before usage; adding a successful $0.12000 Spanish result brings fixed events to $0.548. Set maxTotalChargeUsd for repeated Clip work.
How does Twitch Video Transcript compare with alternatives?
| Option | Best fit | Trade-off |
|---|---|---|
| Twitch Video Transcript | Converting one chosen public highlight into timed spoken text | Requires the exact Clip and does not discover a channel |
| Twitch channel or VOD scraper | Finding broadcasts, schedules, categories, or a creator's archive | Clip-level audio recognition may be secondary |
| Universal transcription Actor | Mixing uploads with media links from many sites | The form is not tuned to Twitch Clip slugs |
| Human transcription | Approving sensitive quotes or resolving chaotic game audio | Adds judgment but takes manual timing work |
Choose this Actor once the public Twitch highlight is known and its speech matters. Video To Text supports uploads and other platforms; Best Video Downloader returns media instead of structured dialogue.
What are the limits and troubleshooting steps?
| Symptom or limit | What it means | Next step |
|---|---|---|
| The Actor rejects the platform | The URL does not resolve as a supported public Twitch Clip | Copy the public clips.twitch.tv/{slug} URL |
| Video information cannot be retrieved | The Clip is expired, deleted, restricted, or temporarily unavailable | Confirm the Clip still plays publicly and retry |
| The media has no usable speech | The Clip is silent, music-only, missing audio, or unintelligible | Select a highlight with clearly audible speech |
Metadata is empty or duration is 0 | Twitch omitted that broadcaster, date, timing, game, or view field | Keep the nullable schema value rather than estimating it |
| Translation is absent | One recognized phrase did not complete target conversion | Retain the source transcript and retry; incomplete translation is not charged |
| A synchronous API call returns HTTP 408 | The client wait ended while the run remained active | Poll asynchronously and retrieve the Dataset after completion |
Segments are ordered phrase windows, not per-word cues. Support reports should omit tokens and private material but include the run ID, a public Clip, expected behavior, and visible error.
Frequently asked questions
Can it process a channel page, live stream, or full VOD?
No. The verified contract is one specific public Twitch Clip URL per run.
Can it crawl a channel's Clip library?
No. The production contract is one public Clip URL per run, not discovery or batch crawling.
Does it use Twitch's existing Clip transcript?
No. Recognition listens to Clip audio; it does not copy a Twitch caption track or return SRT/VTT files.
What happens when a Twitch Clip contains no speech?
A reaction with only music or game audio can yield no detected words. Use a Clip with dialogue for nonempty text.
Why are description, thumbnail, or creator ID empty?
Twitch exposes different fields by Clip. Missing channel, thumbnail, or description values stay nullable instead of being invented.
Is translation billed if one segment fails?
No. The event is emitted only after every target-language segment is present.
Can I use segment timestamps as finished subtitles?
Use them to navigate the highlight or draft captions; precise word alignment and final subtitle formatting need separate review.
Can I schedule the same Twitch Clip URL repeatedly?
Yes. Save the Clip in an Apify schedule, noting that every execution is billed separately and the highlight may later disappear.
Related TrueFetch Actors
- Video To Text processes files and non-Twitch media links.
- Instagram Video Transcript applies a one-post transcript flow to Instagram media.
- Best Video Downloader returns the Clip asset when transcription is not required.
Support and last updated
- Actor page and documentation
- API reference and generated clients
- Issues
- TrueFetch community
- Direct support
Local schema, extractor behavior, metadata pricing, and official documentation links were checked on July 21, 2026. Twitch controls future Clip availability and exposed fields.