Deepfake Detector

Deepfake Detector is a cross-media verification service for journalists, fact-checkers, finance teams, businesses, and individuals. It is strongest when you need both an in-browser check and an API, while its modest free quota suits occasional rather than high-volume work.

Overview

The service analyzes audio, video, and images, returning an Authentic, Likely Synthetic, or Inconclusive verdict alongside a confidence score. That third outcome is useful when evidence does not support a firm call, and the score gives a reviewer a graded signal rather than a simple yes-or-no label. Face analysis is supported, but real-time analysis is not.

Founded in 2023 and headquartered in Montreal, Quebec, Deepfake Detector is aimed at media verification across personal and professional workflows. Its combination of web access, a browser extension, and developer interfaces is more flexible than a single-purpose scanner, although the plan structure makes recurring volume an important buying consideration.

Key features

Video is analyzed frame by frame. Applicable paid plans accept clips up to 10 minutes, while the free tier permits only two minutes per detection, so longer interviews or recordings require a subscription. Stated coverage spans major video generators including Sora, Runway, Pika, and Kling; image generators such as MidJourney, DALL·E, Stable Diffusion, and Flux; and voice platforms including ElevenLabs, PlayHT, Resemble, and OpenAI. This breadth helps when checking varied synthetic-media sources, but a verdict remains inconclusive when the system cannot classify a file confidently.

The browser extension scans image, video, and audio on a webpage from a right-click menu. It runs in Chrome, Edge, Brave, Arc, and other Chromium browsers, making it suitable for checking material in context instead of downloading it into a separate application. The extension says it does not read other tabs, browsing history, DOM content, or cookies, and requests only contextMenus and activeTab permissions.

For automation, the public REST API accepts either a media URL or a file and returns JSON detection results. Type-safe SDKs cover Python, Node, Go, and Ruby. Files longer than 60 seconds can be handled asynchronously through a registered webhook, which is a practical fit for queues and newsroom or compliance pipelines rather than interactive, real-time monitoring.

Uploaded files are processed in memory and purged within 60 seconds after the verdict returns. Hashed identifiers and verdict metadata remain for billing and audit, so organizations with strict retention rules should account for that distinction. The company says its SOC 2 audit is in progress; EU data residency is offered on Business plans and above.

Pricing

The Free plan costs nothing and includes 50 detections per month, web access, AI/Human verdicts, and a TrustScore from 0 to 100. Voice and video are capped at two minutes per detection. It is appropriate for personal checks, occasional fact-checking, or evaluating the workflow, but 50 monthly detections and short clips will constrain regular teams.

Starter costs $4,939 a month, billed $468/year, with 1,000 detections per month. It raises voice and video limits to 10 minutes per detection and adds detailed confidence scores, PDF and CSV export, API access, and email support. Those exports and integrations suit a small operational team, though the stated billing figures should be checked carefully before committing.

Business costs $199,159 a month, billed $1,908/year, for 5,000 detections monthly. It retains the 10-minute media limits and API access, and upgrades to priority email support. Enterprise costs $599,479 a month, billed $5,748/year, and increases the allowance to 20,000 detections with the same 10-minute voice and video limits, API access, and priority support. Business and Enterprise are the relevant tiers for heavier organizational workloads and EU residency.

Platforms

Deepfake Detector is offered through the web, a public API, and a browser extension. The extension supports Chrome, Edge, Brave, Arc, and other Chromium browsers. SDKs are available for Python, Node, Go, and Ruby, giving engineering teams several maintained language entry points.

Who it's for

Choose it for journalism, fact-checking, finance, business investigations, or personal verification when audio, video, and image coverage matter together. The free plan works for light use; Starter fits teams that need exports and automation; Business and Enterprise fit higher monthly volumes and organizations requiring priority support or EU data residency. It is not the right choice for real-time detection or workflows that cannot retain hashed identifiers and verdict metadata.

Pros and cons

Pro: One service covers audio, video, and images with a three-way verdict and confidence scoring.

Pro: Frame-level video analysis, 10-minute paid limits, webhooks, REST access, and four SDKs support production workflows.

Pro: The extension offers in-page checks with narrowly stated browser permissions.

Con: The free allowance is only 50 detections monthly, with two-minute voice and video caps.

Con: There is no real-time analysis, and retained hashes and verdict metadata may conflict with some data policies.

Con: The published monthly and annual billing figures for paid plans warrant confirmation before purchase.

Alternatives

For a broader directory of choices, see AI Deepfake Detection Tools. Pick Netarx instead if you need mobile and desktop apps alongside an extension and web access, with free Individual protections. ForgeSpy is a freemium web and mobile option built around sign-up credits for image and video scans. Choose Unreveal when an API, Linux, macOS, Windows, and web deployment matter more than a free tier. FauxLens offers a 40-credit free plan, four detections, and a six-signal forensic pipeline across Android, API, extension, and web. Hive AI-Generated Content Detection suits developers wanting access to more than 10 models with pay-as-you-go credits. Reality Defender provides a 50-scan monthly free API plan with up to three seats and image/audio analysis. Sensity AI is a paid service with a free trial and self-hosted support. DuckDuckGoose DeepDetector is another paid option with free access and a trial.

Verdict

Deepfake Detector is a strong fit for teams that need one detector for images, audio, and video plus browser and API access. Its main reason to choose it is the combination of broad generator coverage, asynchronous integrations, and clear verdict categories. Look elsewhere if you need real-time analysis, larger free usage, or retention terms that exclude metadata after processing.