> ## Documentation Index
> Fetch the complete documentation index at: https://composio-preview.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# TruthScan

> TruthScan detects AI-generated and manipulated content in images, text, PDFs, video, and audio.

![TruthScan logo](https://mintcdn.com/composio-preview/IA0_SoMcsWt17HEC/images/toolkits/truthscan.svg?fit=max\&auto=format\&n=IA0_SoMcsWt17HEC\&q=85\&s=586f093ce805c1cabe9f911998c6905d)

* **Category:** artificial intelligence
* **Auth:** API\_KEY
* **Composio-managed OAuth available?** N/A
* **Tools:** 8
* **Triggers:** 0
* **Slug:** `TRUTHSCAN`
* **Version:** 20261101\_00

## Tools

> [accordion] Analyze Audio
>
> **Slug:** `TRUTHSCAN_ANALYZE_AUDIO`
>
> Submit exactly one public audio URL or uploaded audio/media file for AI-generation detection. By default, analyze up to the first 60 seconds and wait with bounded polling for the latest result; submission consumes account credits.
>
> #### Input Parameters
>
> **audio\_url** `string`
>
> Publicly reachable audio or supported media URL to analyze.
>
> **audio\_file** `object`
>
> Audio or media attachment to upload directly. Supported extensions are mp3, wav, m4a, flac, ogg, and mp4; the provider accepts files from 1 KB through 100 MB.
>
> **timeout\_seconds** `number`
>
> Maximum total polling time in seconds (1-600). On timeout, the latest nonterminal status and job ID are returned for later resumption.
>
> **wait\_for\_completion** `boolean`
>
> Poll until the analysis is done or failed. Set false to return the accepted job immediately.
>
> **analyze\_up\_to\_seconds** `integer`
>
> Maximum leading duration to analyze, in seconds.
>
> **poll\_interval\_seconds** `number`
>
> Seconds between result checks while waiting (1-60).
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Analyze Image
>
> **Slug:** `TRUTHSCAN_ANALYZE_IMAGE`
>
> Submit an image URL for AI-generation detection and, by default, wait for the asynchronous result. Returns the job ID when not waiting or when polling times out; detection consumes account credits.
>
> #### Input Parameters
>
> **model** `string`
>
> TruthScan model name; use generic unless a specific instance/model is known.
>
> **image\_url** `string` *required*
>
> Publicly reachable image URL (1 KB-10 MB) to analyze.
>
> **timeout\_seconds** `number`
>
> Maximum total wait time in seconds (up to 600); on timeout return the current pending status and job ID for later resumption.
>
> **generate\_heatmap** `boolean`
>
> Generate a secured detection heatmap.
>
> **generate\_preview** `boolean`
>
> Generate a secured preview asset for later download.
>
> **wait\_for\_completion** `boolean`
>
> Poll until done or failed; set false to return immediately with the job ID.
>
> **poll\_interval\_seconds** `number`
>
> Seconds between result checks when waiting (minimum 1).
>
> **generate\_analysis\_details** `boolean`
>
> Include detailed analysis indicators in the completed result.
>
> **generate\_heatmap\_overlayed** `boolean`
>
> Override generation of the overlay heatmap variant; omit to use the provider default, and do not set true when generate\_heatmap is false.
>
> **include\_content\_moderation** `boolean`
>
> Include content-moderation signals in the analysis.
>
> **generate\_heatmap\_normalized** `boolean`
>
> Override generation of the normalized heatmap variant; omit to use the provider default, and do not set true when generate\_heatmap is false.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Analyze PDF
>
> **Slug:** `TRUTHSCAN_ANALYZE_PDF`
>
> Submit a publicly reachable HTTPS PDF URL for asynchronous tampering detection via the text detector host and, by default, wait for a result. Each page costs 1,000 credits; returned fields omit the provider's PDF summary and modules.
>
> #### Input Parameters
>
> **model** `string ("pdf_detector" | "pdf_detector/v1" | "pdf_detector/v3" | "pdf_detector/v4")`
>
> PDF detector version. Use pdf\_detector/v4 for current tampering detection; v1 focuses on AI-generation fingerprints and v3 on structural tampering.
>
> **pdf\_url** `string` *required*
>
> Publicly reachable HTTPS URL ending in .pdf. The PDF must be no larger than 2 MB; analysis costs 1,000 credits per page.
>
> **timeout\_seconds** `number`
>
> Maximum total wait time in seconds. On timeout, returns the latest pending status and document ID so the job can be resumed later.
>
> **wait\_for\_completion** `boolean`
>
> Poll until done or failed; set false to return immediately with the document ID.
>
> **poll\_interval\_seconds** `number`
>
> Seconds between result checks when waiting (minimum 1).
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Analyze Text
>
> **Slug:** `TRUTHSCAN_ANALYZE_TEXT`
>
> Submit a text value or public web URL for AI-written-text detection. Provide one input source and never both. By default, wait for the asynchronous result; detection consumes account credits.
>
> #### Input Parameters
>
> **url** `string`
>
> Publicly reachable web URL whose text should be analyzed. Required when text is omitted; do not provide both fields.
>
> **text** `string`
>
> Text to analyze, under 30,000 words; at least 200 words is recommended. Required when url is omitted; do not provide both fields.
>
> **model** `string`
>
> TruthScan text detector model.
>
> **timeout\_seconds** `number`
>
> Maximum total wait in seconds (1-300). On timeout, return the latest pending status and document ID for later resumption.
>
> **wait\_for\_completion** `boolean`
>
> Poll until the detection is done or failed. Set false to return immediately with the document ID.
>
> **poll\_interval\_seconds** `number`
>
> Seconds between result checks when waiting (1-60).
>
> **generate\_analysis\_details** `boolean`
>
> Include deeper source and analysis signals when supported.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Analyze Video
>
> **Slug:** `TRUTHSCAN_ANALYZE_VIDEO`
>
> Analyze exactly one public video URL or uploaded video for AI generation or face swapping. By default, wait with bounded polling and return the latest result; this non-idempotent submission consumes account credits.
>
> #### Input Parameters
>
> **model** `string ("generic" | "faceswap")`
>
> Detection model: generic detects AI-generated video, while faceswap detects face swapping.
>
> **video\_url** `string`
>
> Publicly reachable video URL to analyze.
>
> **video\_file** `object`
>
> Video attachment to upload directly. Supported extensions are mp4, mov, avi, mkv, and webm; the provider accepts files from 1 KB through 100 MB.
>
> **timeout\_seconds** `number`
>
> Maximum total polling time in seconds (1-600). On timeout, the latest nonterminal status and job ID are returned for later resumption.
>
> **generate\_preview** `boolean`
>
> Generate a preview when the selected detector supports one.
>
> **wait\_for\_completion** `boolean`
>
> Poll until the analysis is done or failed. Set false to return the accepted job immediately.
>
> **poll\_interval\_seconds** `number`
>
> Seconds between result checks while waiting (4-60). Video has a low request limit, so shorter intervals are rejected.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Check Credits
>
> **Slug:** `TRUTHSCAN_CHECK_CREDITS`
>
> Return the connected TruthScan account's remaining detection credits from the canonical image detector balance endpoint.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Download Image Asset
>
> **Slug:** `TRUTHSCAN_DOWNLOAD_IMAGE_ASSET`
>
> Download a generated preview or heatmap for an owned TruthScan image detection job. Returns a managed file when ready, or an explicit pending or unavailable state.
>
> #### Input Parameters
>
> **job\_id** `string` *required*
>
> Sensitive image document/job ID returned by Analyze Image.
>
> **asset\_type** `string ("preview" | "heatmap")` *required*
>
> Generated image asset to download: preview or heatmap.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Detection Result
>
> **Slug:** `TRUTHSCAN_GET_DETECTION_RESULT`
>
> Get the current status and available results of an existing image, text, PDF, video, or audio detection job without submitting another analysis. Optionally poll until the job finishes or the wait times out. PDF results do not include the provider's summary or module details.
>
> #### Input Parameters
>
> **job\_id** `string` *required*
>
> Sensitive document or job ID returned by a TruthScan analysis tool.
>
> **modality** `string ("image" | "text" | "pdf" | "video" | "audio")` *required*
>
> Detector that created the job. Text and PDF share a host but retain distinct result semantics.
>
> **timeout\_seconds** `number`
>
> Maximum total wait in seconds (1-300).
>
> **wait\_for\_completion** `boolean`
>
> Poll until done or failed instead of returning the current status once.
>
> **poll\_interval\_seconds** `number`
>
> Seconds between checks when waiting (1-60). Video requires at least 4 seconds because its query limit is lower.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not
