Code your transcripts in minutes. Every quote checked. Every decision yours.

For theses, dissertations and research projects.

Frameworks:Braun & ClarkeThematic Synthesis+ 3 more
Code my transcripts freeSee how it works
3 sources free, no signupEvery quote checked against your textYour data never trains an AI

Project · Nurses' wellbeing after night shifts

Example

4

participants

20

codes

3

themes

▾ Carrying the shift homeMoral distress at the bedsidePeers as a lifeline

Traced to Participant 1

Interviewer: What happens after a difficult shift? Participant 1: I sit in the car park for twenty minutes before I can drive. I leave the ward, but the ward doesn't leave me. I replay every alarm when I try to sleep.

Code: carrying the shift home

✓ Quote found word for word in Participant 1

Distress from the shift follows the nurse into rest and sleep.

Theme: Carrying the shift home

3 of 4 participants
✎ Rename✎ Move code✎ Remove✎ Add your own

Built for how you work

  • Students

    Weeks of coding before your deadline, and NVivo needs a course to learn.

    Upload your transcripts and get every one coded in minutes, then refine the codes and themes yourself. Nothing to install or learn first.

    Code my transcripts →

  • Academics and research teams

    You need speed without losing a method you can defend to reviewers.

    A named framework, verbatim evidence for every code, your own codebook if you have one, a methods paragraph and an audit trail. Export to NVivo when you want to.

    Try it on my data →

  • Literature reviews

    Synthesising 20–50 papers by hand takes months.

    Upload the PDFs. Findings are coded study by study, and every theme lists the studies behind it, ready for a thematic synthesis.

    Synthesise my studies →
● 3 steps

How it works

From raw papers or transcripts to themes you can check and cite.

Step 01

Add your data

Upload PDFs, Word files, transcripts or a survey spreadsheet, and say whether it's published studies or your own participants.

  • PDF · Word · TXT · VTT
  • CSV · Excel
Step 02

Pick a method and run

Choose one of 5 frameworks. Each source is coded on its own, then the codes are grouped into themes across all of them.

  • Braun & Clarke by default
  • Quotes checked word for word
Step 03

You review, then export

Nothing is final until you say so. Check any quote in its source, rename themes, move or remove codes, add your own, then download your report or open the project in NVivo.

  • Word report (free)
  • Figure 1 map · NVivo · CSV
See the check

AI proposes. Your text decides.

Below, the AI proposed four codes from one interview. Each quote is searched for in the transcript. Two are in it. Two are not, and those codes are removed. Tap a code to see why.

Your transcript · Participant 3 (example)

I had been on the ward for three nights and honestly I felt unsafe. There were never enough staff, and when I pressed the call button nobody came for forty minutes. The night nurse was lovely, she sat with me and held my hand, but she was on her own with twenty patients.

Nothing in the transcript matches this quote, so nothing is highlighted.

Codes the AI proposed

2 kept · 2 removed

Audit trail: the 2 removed codes are listed in your report with the reason, quote not found in source. A general chatbot has no such check.

Quotes in the write-up are checked too

Participant 3 recalled that “nobody came for forty minutes”, and that “the night nurse was on her own with twenty patients”.

The first quote is in the transcript above, word for word. The second was reworded (the text says “she was on her own with twenty patients”), so it loses its quotation marks and is listed in your report. A quote credited to the wrong participant is corrected the same way.

Only the participant’s own words count

Interviewer: So you felt cut off from your team the whole time?

Participant 5: Exactly that. Nobody checked in on me for weeks.

  • Isolation “you felt cut off from your team” ✕ The interviewer's words, not the participant's.
  • Feeling unsupported “Exactly that.” ✕ Only agrees with the question. The meaning is the interviewer's.
  • Unnoticed by colleagues “Nobody checked in on me for weeks.” ✓ Kept
Run this check on your own transcript →

Example excerpt, invented for this page. The same check runs on your text.

What it can do for you

Pick the kind of data you have.

For transcripts, focus groups and open-ended answers: 3 participants free, up to 50 with a pass.

  • Free

    Coding participant by participant

    Each transcript is coded separately with quotes. Files holding several interviews are split for you.

    Learn more →
  • Free

    Themes and who said what

    Themes across your dataset, a participant matrix and a reflexivity memo on each theme.

  • Advanced pass

    Table 1: participant details

    Age, role and other details each participant states, pulled into a table. Nothing guessed.

  • Advanced pass

    Compare groups

    How often each theme comes up per group, such as nurses vs doctors, with a quote from each.

  • Advanced pass

    Saturation curve

    New codes per participant, whether saturation was reached, and a methods paragraph to paste.

  • Advanced pass

    Ask your data

    Ask what participants said about a topic and get answers from your transcripts only.

Try it free with your transcripts →

Every analysis exports to

Map downloads, CSV and .qdpx need a pass from $9. Compare plans

We keep no copy

Your data is analysed and sent straight back. Nothing is saved to a database or a log.

Your data never trains an AI

Your participants' words are never added to an AI model, so they can never resurface in anyone else's results.

No account needed

Nothing to sign up for. Your results live in this browser and are gone when you close the tab.

Writing an ethics application? The privacy policy sets out exactly what is processed and what is not.

🛡️ Why not just use a chatbot?

Built for work you have to defend

A chatbot summary is hard to cite and harder to defend in a viva. Here is the difference.

A method you can cite

✕ General chatbot:A summary with no named method behind it.
✓ thematicanalysis.ai:Follows the framework you pick, such as Braun & Clarke or Thomas & Harden, and names it in your report.

Quotes you can check

✕ General chatbot:Quotes can be reworded or invented.
✓ thematicanalysis.ai:Every quote, including those in the written analysis, is checked word for word against your text. Ones that fail are removed and listed.

Who said what

✕ General chatbot:Blends every source into one summary.
✓ thematicanalysis.ai:Codes each study or participant separately, so every theme lists the sources behind it.

Ready for your thesis

✕ General chatbot:Chat text you reformat by hand.
✓ thematicanalysis.ai:A Word report with themes, quotes, limitations and an audit trail, plus NVivo export.

Transparent Pricing

Simple, transparent academic pricing

Start completely free with your first 3 sources. Pay once for 30 days of full access — re-run and revise your analysis as often as your pass allows, with nothing to cancel.

Starter

Free forever

Free

/ no card needed

Test the analysis workflow on small samples before committing your full project.

Sample trial includes:

  • ✓Up to 3 sample sources or participants
  • ✓2 test analysis runs (initial + 1 re-run)
  • ✓Full-length sources: whole papers and transcripts
  • ✓Literature Review & Primary Research modes
  • ✓All 5 qualitative frameworks (Braun & Clarke, etc.)
  • ✓Verbatim quote evidence & source matching
  • ✓Standard Word (.docx) & Markdown export
Try 3 sources free

No signup or credit card required

Popular

Thesis Sprint Pass

30-Day Pass

$9

/ one-time payment

Full project capacity for completing your dissertation, thesis, or academic paper.

Everything in Starter, plus:

  • ✓Analyse up to 20 full sources or transcripts (analyze your entire study)
  • ✓Up to 5 full analyses / re-syntheses across 30 days
  • ✓Batch multi-file PDF, DOCX, CSV & Excel dropzone (up to 10MB per file)
  • ✓High-Res Figure 1 Thematic Maps (300DPI PNG & Vector SVG for publication)
  • ✓Universal REFI-QDA (.qdpx) export for NVivo, ATLAS.ti, MAXQDA & Dedoose
  • ✓Structured CSV / Excel Codebooks & Cross-Tabulation Matrices
  • ✓30 days full project access — revisit, tweak, and re-export anytime

30 days access · One payment of $9 · No subscription

Advanced Project Pass

30-Day Pass

$15

/ one-time payment

Extended access and advanced analytical tools for complex, deeper qualitative projects.

Everything in Thesis Sprint, plus:

  • ✓Analyse & synthesise up to 50 full sources or studies per project, with 25 extra swaps to fix or replace sources
  • ✓Participant Demographics Table (Auto-extracted, 3 generations)
  • ✓Theme Comparison by Participant Group
  • ✓Code Saturation Curve with a ready-to-paste methods paragraph (Guest et al., 2020)
  • ✓Research Gap Finder & Theoretical Underpinnings for Literature Reviews (3 generations each)
  • ✓Ask Your Data (Q&A Chat) (Up to 7 queries grounded in data)
  • ✓Extended 30 days full project access
  • ✓Up to 10 full analyses / re-syntheses

30 days access · One payment of $15 · No subscription

Detailed Matrix

Compare Plan Features

Every tier includes strict data privacy, verbatim grounding, and academic export.

FeatureStarterThesis Sprint PassAdvanced Project Pass
Project Scale
Max Sources / Participants32050
Full Analysis Runs2510
File Uploads (PDF, DOCX, CSV, Excel)✓ Up to 3MB✓ Up to 10MB✓ Up to 10MB
Core Analysis Engine
Thematic Frameworks & Verbatim Grounding✓✓✓
High-Res Thematic Maps (PNG/SVG)Preview✓ Included✓ 3 generations each
Divergence & Agreement MatrixBasicFull (20 sources)Full (50 sources)
Advanced Tools
Participant Demographics Table——✓ 3 generations
Theme Comparison by Participant Group——✓ Unlimited
Code Saturation Curve & Methods Text——✓ Unlimited
Research Gap Finder & Theory——✓ Included
Ask Your Data (Q&A Chat)——✓ 7 queries
Export & Access
Word, CSV, Excel & REFI-QDA ExportWord only✓ Full Suite✓ Full Suite
Access DurationSession-based30 Days Access30 Days Access
🔒100% In-Browser Privacy

Uploaded PDFs and pasted transcripts are parsed locally in your browser. Your raw data is never used to train public models.

🎓Methods Chapter Ready

Every report generates a complete audit trail with SHA-256 data hashes, quote verification records, and reflexivity prompts for viva defense.

🚫One-Time Payment, No Hidden Fees

Paid project passes are a one-time payment for 30 days of access. No recurring billing, no surprise renewals.

We keep no copy

Your data is analysed and sent straight back. Nothing is saved to a database or a log.

Your data never trains an AI

Your participants' words are never added to an AI model, so they can never resurface in anyone else's results.

No account needed

Nothing to sign up for. Your results live in this browser and are gone when you close the tab.

Writing an ethics application? The privacy policy sets out exactly what is processed and what is not.

Your first 3 sources are on us

Upload 3 studies or transcripts, pick a framework and see your themes. No signup, no card.

Start your free analysis

FAQ

Frequently asked questions

Yes, that is what it is built for. Use literature review mode for your review chapter and interview mode for your own participant data. You get codes with a quote behind each one, themes with the sources that support and contradict them, and a Word report to work from. Treat it as a first pass: review and rename the codes yourself, cite the framework you chose (Braun & Clarke, Thomas & Harden and so on) as your method, and disclose AI assistance where your university requires it.

Your first 3 sources are free, with 2 runs, no signup and no card. For more, the $9 Thesis Sprint Pass covers up to 20 sources and 5 runs, and the $15 Advanced Project Pass covers up to 50 sources and 10 runs plus the Advanced tools. Both are one-time payments for 30 days, with no subscription. Every tier reads full-length sources, whole papers and transcripts. A run that fails is not counted.

Up to 50 sources and 10 runs, plus tools that appear under your results. For literature reviews: the Research Gap Finder and Theoretical Underpinnings. For interviews and surveys: a participant demographics table (Table 1), theme comparison by participant group, and a code saturation curve with a methods paragraph (Guest, Namey & Chen, 2020). Both modes get Ask Your Data chat (7 questions). The gap finder, theory and demographics tools include 3 generations each; the group comparison and saturation curve are calculated from your results and are unlimited.

For a literature review, upload each paper as a PDF or Word file. The tool keeps the methods, findings, discussion and conclusion, reads the authors and year for the citation label, and leaves out the rest, so you do not need to trim anything. For your own research, upload interview or focus group transcripts (Word, PDF, TXT or VTT) or survey exports (CSV or Excel). A file holding several participants is split into one each. If a file cannot be read, such as a scanned PDF, you are told which one and how to fix it.

Every code carries a quote, and every quote is checked word for word against the text you uploaded. A code whose quote cannot be found is removed and listed in the audit trail at the end of your report. Quotes the AI uses inside its written analysis are checked the same way, and an interviewer's question or a bare "yes" is never counted as a participant's evidence. Each theme lists the sources and codes behind it, and you can click any code to see its quote in the original document.

Five: Reflexive Thematic Analysis (Braun & Clarke), the default; Thematic Synthesis (Thomas & Harden) for literature reviews; Framework Analysis (Ritchie & Spencer); Grounded Theory with constant comparison; and Qualitative Content Analysis. The analysis follows the terminology of the framework you pick, and the report names it so you can cite it.

For most student projects, yes. A student NVivo licence costs about $125 a year, and more with the AI add-on. Here 3 sources are free and a whole project costs $9 or $15 once, with nothing to install. You can also export to NVivo, MAXQDA or ATLAS.ti. Check your university library first, since a site licence may make NVivo free for you, and for a large funded study with hundreds of transcripts NVivo or ATLAS.ti is still the better fit.

Your text is sent to our AI provider to produce your results and is not stored on our servers. We use providers that do not train their models on API data. Your project is saved in your own browser on this device, and you can download the report or a .ta project file to keep a copy. Remove names and identifying details from participant data before uploading.

Yes, within 7 days of buying a pass, if an analysis fails and still fails when you retry. Email hello@thematicanalysis.ai with the email you paid with, the pass and the error you saw. Refunds are not given for completed analyses, files that could not be imported or a change of mind. Full details are on the Refund Policy page.

No. Open the tool, add your data, run the analysis and download the report. There is nothing to sign up for or cancel.