Answers before you upload
Frequently Asked Questions
Everything you need to know about analyzing your WhatsApp conversations with ThreadRecap.
- Photos stay on your device; from a video, only the audio track your browser extracts is sent.
Frequently Asked Questions
Privacy & Data
Yes. Your chat files are unzipped and parsed locally in your browser. Photos and videos never leave your device. The text and the audio of every voice message are sent to our secure servers (encrypted in transit). Your chat is stored on our servers so your analysis stays accessible. You keep full control: delete your chat and any analysis anytime from your dashboard. Never used for ads or training and shared only with the providers that run the service. From a video, only the audio track your browser extracts is sent for transcription; the video file and its pictures stay on your device.
No. Photos and videos stay on your device — the files are never uploaded. We do recognize where they appear in the conversation, so your record shows that a photo or video was sent at that point. Chat text is sent, and so is the audio of every voice message. From a video, only the audio track your browser extracts is sent for transcription; the video file and its pictures stay on your device.
Your chat is stored on our servers so your analysis stays accessible. You keep full control: delete your chat and any analysis anytime from your dashboard. Never used for training and shared only with the providers that run the service. See our Privacy Policy for details.
Audio files and transcripts can be saved with your analysis so you can replay voice messages and revisit the conversation without uploading it again. You can delete the saved audio and transcripts while keeping the analysis, or delete the entire analysis, at any time.
Getting Started
Open WhatsApp, open the chat, open its menu (the chat name on an iPhone, or the three dots at the top right in a group, the three dots then More on Android), and choose Export Chat. Pick "Include media" if you want the voice messages transcribed. WhatsApp then asks where to put the file: Save to Files on an iPhone, Google Drive on Android. Some Android phones also show Save as file, which puts it in Downloads. That last step is not optional on Android, because the file is not stored on the phone until you choose.
Yes. We support the standard WhatsApp export: a .zip that contains a _chat.txt file. You can also use a .txt-only export for text-only analysis. Voice messages inside the .zip are supported when you include media.
OPUS is an open audio codec that WhatsApp uses for voice messages. When you export a chat with media, voice messages often appear as .opus files. We detect and transcribe them automatically — no conversion needed.
Parsing is instant in your browser. Most conversations are ready within a few minutes. Conversations with many voice messages take longer.
Analysis & Summaries
The upload screen shows the conversation's overview at no cost: who talks, when, how many voice messages, and one of them already written out. It also shows the exact price to unlock the full conversation. You only pay if you want to.
An unlock gives access to the full conversation, a chat allowance, and one document of your choice. You can then buy another document or add more chat use for one fixed price.
No. Upload any chat with decisions, action items, or open questions. After the initial recap, ask the follow-up chat for meeting minutes and it will organize the agenda, decisions, actions, and unresolved points.
Yes. Ask the follow-up chat for every question that was asked but never answered, or every topic left for later, and it lists them with citations to the original messages.
We use who said what: "I'll send the deck" → that person owns it. "Can you review by Friday?" → the person being asked. If it's ambiguous (e.g. "someone should…"), we label it "Unassigned." Deadlines are only added when clearly stated.
The interface is available in 6 languages. Each voice message is written in the language it was spoken, and you choose the language of the analysis.
Wrapped-style tools focus on fun statistics and shareable cards. ThreadRecap is built for substantive analysis: timelines, exact quotes, decision logs, voice-message transcription, action items, meeting minutes, and dispute documentation.
Voice & Audio
We support OPUS, OGG, M4A, MP3, AAC, and WAV when they appear in your export. Audio is processed through OpenAI's audio transcription API and merged into your conversation timeline.
Transcription quality varies with recording quality, accent, background noise, and overlapping voices. Read the transcript against the original recording before quoting it.
Yes. Voice messages in your export are transcribed and merged into the timeline before analysis. So "we agreed to launch on March 15" in a voice message is included in decisions and actions.
Today we process all voice messages in the export you upload. WhatsApp cannot export only some messages, so to transcribe a subset, use the date range filter after uploading.
Not as a subtitle file. Subtitle cues are timed against one recording, and a WhatsApp export is a conversation — dozens of separate messages from more than one person. What you can download instead is the whole conversation as a .txt or a PDF, every line stamped with its date and time and every voice message written out in the place it was sent.
Groups & Participants
Yes. You can select which participants to include in the reading, and the upload screen shows the exact price before you pay. The unlock is charged once for the conversation, based on its size and voice messages — with no group multiplier.
Yes. Focus the analysis on 2–5 participants. Messages from everyone else provide context but the AI prioritizes what your chosen people said. That cuts most of the off-topic noise while keeping decisions and action items.
Export without media to stay within WhatsApp's limits and reduce file size. Focus on a subset of participants to keep the recap tight. For very large groups, pick the people who usually drive decisions and follow-ups.