Tayanch

Tayanch is a media-analysis service for checking whether images and videos are real or AI-generated. It suits developers and product teams that want API-based checks, including a free allowance for evaluation. Its low per-image price is attractive, but its verdicts are probabilistic and audio analysis is not currently available.

Overview

Tayanch combines several kinds of evidence instead of forcing a binary answer: ambiguous or heavily processed media can receive an uncertain verdict. It checks Content Credentials and generator markers, alongside vision-transformer embeddings, camera physics and compression. That mix can identify fully generated images, localize likely AI edits in real photos with a heat map, and flag face manipulation or generated video at the frame level. Probability and confidence make the result more informative, but they do not make it proof.

The API uses REST over HTTPS, returns JSON and authenticates with an API key in the X-API-Key header. It accepts files and media links; the URL endpoint fetches media, and social previews resolve automatically. A web interface is also available. Tayanch says submitted media is sealed on the client using TLS and an HPKE envelope, analyzed in memory, discarded after the verdict and never used for training. It says it stores only a cryptographic hash of each API key and does not retain submitted media.

Audio analysis for voice clones and synthetic speech is under maintenance and coming soon, so Tayanch is not yet a choice for teams needing a complete image, video and audio workflow. The website, demo, free tier and self-serve API have no SLA; an SLA is available only under a signed enterprise agreement. Tayanch cautions against using its verdict as the sole basis for accusations or legal, customer, content-removal or disciplinary action.

Key features

• Image checks: Detect fully generated images and show a heat map to localize AI edits in a real photo. This is useful when a reviewer needs to see where suspected alteration may have occurred, not just receive a label.
• Video checks: Identify face manipulation and provide frame-level advisories for generated footage. The frame-level signal adds context, but remains advisory rather than conclusive.
• Uncertain verdicts: Ambiguous and heavily processed inputs can be marked uncertain instead of receiving a forced guess. That restraint is preferable for cautious screening, though it can leave users without a decisive answer.
• API and link handling: REST, JSON, API-key authentication and support for files and media links make Tayanch suitable for products that need to integrate checks. Automatic social-preview resolution and URL fetching reduce the need to handle media manually.
• Deployment and support: On-premise deployment is available by contacting Tayanch. Enterprise plans include dedicated thresholds, SLAs and support, which are the relevant options for organizations that need contractual service commitments.

Pricing

Tayanch is freemium, with a no-card API allowance and per-check options for teams that do not want a monthly commitment. The free quota is 100 checks and permits 30 requests per minute; it is useful for evaluation, but the 100-check cap makes it a limited basis for sustained production use.

The Team plan's 50,000 checks suit production workloads, but its monthly price is a substantial step up from pay-as-you-go for smaller or uneven volumes. Enterprise is the route to an SLA; the free tier and self-serve plans do not provide one. No free trial is offered.

Platforms

Tayanch is available on the web and through an API. Its image and video checks support media types beyond a web-only workflow, but audio analysis is not currently available. On-premise deployment can be arranged through Tayanch.

Who it's for

Tayanch is a good fit for developers who need image or video analysis in an application and want to begin with a no-card API allowance before paying per check or moving to a production plan. Teams handling sensitive media may value its stated handling and retention practices. It is a weaker fit for audio-first workflows, organizations that require an SLA on a self-serve plan, or anyone who needs detection verdicts to make consequential decisions on their own.

Pros and cons

• Pros: The 100-check free API quota requires no card, making evaluation possible without an upfront commitment.
• Pros: Image checks can localize AI edits with a heat map, while video checks provide frame-level advisories rather than only a single overall label.
• Pros: An uncertain verdict gives users a way to avoid treating ambiguous media as definitively real or generated.
• Pros: Per-check image and video pricing avoids a monthly commitment, and on-premise deployment is an option for customers who need it.
• Cons: Audio analysis is under maintenance, so voice-clone and synthetic-speech checks cannot yet be part of the workflow.
• Cons: The free allowance stops at 100 checks, while production use is positioned on a $249 a month Team plan.
• Cons: The self-serve service has no SLA, and verdicts are expressly unsuitable as the sole basis for consequential action.

Alternatives

AI Deepfake Detection Tools is a useful starting point for comparing options across the category.

• Netarx is worth considering instead for someone seeking individual video, voice and email protection across mobile, desktop and browser platforms, with a free Individual plan.
• FauxLens is an alternative for users who want Android or browser access and a free allowance of 40 credits across four detections, with a six-signal forensic pipeline.
• ForgeSpy is another option for mobile users, with Android and iOS apps and five free signup credits.
• Deepfake Detector may suit users looking for an API, browser extension and web service, with a free plan offering 50 detections per month and up to two minutes per voice or video detection.
• Hive AI-Generated Content Detection is an alternative for teams seeking API, extension, web or self-hosted access and access to more than ten Hive models through its developer plan.
• Reality Defender may be preferable for a free plan that explicitly includes image and audio analysis, 50 scans per month and up to three seats.
• Resemble Detect is an option for teams wanting audio, image and video detection across platform and API access, with a free Flex plan and pay-as-you-go credits.
• Unreveal is another alternative.

Verdict

Choose Tayanch if you need integrated image or video checks, value an uncertain result over a forced guess, and want to start with a no-card API quota or pay per check. Its main advantage is a combination of practical media signals, including image-edit localization and frame-level video advisories. Look elsewhere if audio checks are essential, a self-serve SLA is required, or you need a verdict that can stand alone in a consequential decision.