Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
The deepfake detector is a tool developed within the WeVerify project that analyzes images or videos to return the probability that the media contains deepfake-manipulated faces. For videos, it segments the footage into shots and derives frame-level probabilities, with the final video probability taken from the highest-shot probability; input is a URL of the suspicious media and output is a deepfake probability score. It is available as a standalone demo and via a REST API for programmatic use, with trial requests via [email protected], and is funded by the EU Horizon 2020 WeVerify project (grant No 825297).
Parse Score