Analyze the Full Scope of Legal Evidence, Not Just the Documents

Legal work has never lived entirely in text — exhibit images, deposition recordings, org charts buried in a merger agreement. Until now, Harvey could read the words around these but not the visuals or recordings themselves, which meant pulling files out to review by hand, or sending audio to a transcription vendor and waiting to re-upload it before analysis could even start.

That’s why Harvey now supports audio and video transcription along with enhanced image understanding.

Enhanced Image Understanding

Harvey can now visually interpret images, charts, graphs, and diagrams. Upload a standalone image, or a PDF, Word, or PowerPoint file with visual content, and ask your question. Then, Harvey will spot when visual inspection is needed and analyze the relevant page before answering.

Deal teams can read ownership structures straight out of a purchase agreement without redrawing them. Litigators can review patent drawings or trademark specimens directly. In-house teams working through chart-heavy filings or board materials can just ask Harvey what a chart shows, instead of transcribing it into text first.

Audio File Transcription

Drop an audio file into Assistant, or upload it to Vault, and Harvey transcribes it in the background — speaker-labeled, timestamped, and queryable like any other document. Supported formats: M4A, MP3, WAV, WebM, FLAC, OGG. Up to 500MB in Assistant, 4GB in Vault, two hours per file.

You can also record directly from the Harvey mobile app for depositions or client meetings for up to two hours, and the transcript saves straight to Vault with each speaker automatically labeled. Only the transcript is retained, not the original audio.

For law firms: Client calls, witness interviews, and internal case discussions can go straight from recording to searchable transcript, with no separate transcription step and no manual cleanup before analysis can start.

For in-house legal teams: Recorded regulatory interviews, internal investigation interviews, or board and compliance calls become queryable Harvey documents in minutes rather than sitting in an audio file that has to be reviewed in real time.

Video File Transcription

The same transcription works for video: MP4, MOV, AVI, DAV, WebM, up to two hours per file. For litigation teams, this means bulk-uploading a set of deposition videos into Vault and querying across all of them at once: surfacing contradictions, tracing a timeline, pulling every mention of an issue into a Review Table.

For law firms: This is built for the reality of litigation, where video evidence is constant: depositions, witness interviews, surveillance footage. A team can bulk-upload dozens or even hundreds of deposition videos into Vault and query across the entire set at once surfacing contradictions between witnesses, tracing a timeline across statements, or pulling every reference to a specific issue into a Review Table.

For in-house legal teams: Recorded interviews and video evidence sit alongside contracts and correspondence, and native video support means all of it can be reviewed in one platform instead of split across tools.

Why This Matters

With image, audio, and video now natively understood, Harvey can support end-to-end analysis, research, and drafting across the full range of evidence a matter actually contains, not just the parts that happen to be text. That means fewer handoffs between tools, faster turnaround from raw file to usable answer, and a more complete view of the record. Legal work has always been multimodal. Now Harvey is too.

Ready to see it in practice? Contact your account team or reach out to us below to get started.