Qualitative analysis
with AI you can audit.
QualiLens runs five established qualitative methods — from grounded theory to literature synthesis — with LLM coding that pauses for your review, keeps verbatim provenance on all excerpts, and stays entirely on your machine.
The workflow
Five steps to a complete analysis
A five-step wizard handles method selection, configuration, model choice, data upload, and cost estimation. The pipeline then runs, pausing at each checkpoint for your review.
Choose a method
Grounded theory, thematic analysis, content analysis, framework coding, or literature synthesis.
Configure
Set your research question, method-specific options, and any codebook or framework.
Pick a model
Anthropic, OpenAI, Google, or Mistral — your key, your account, direct connection.
Upload data
Drag in documents, PDFs, or recordings. Transcription runs automatically.
Review & run
See a cost estimate, start the pipeline, and approve each checkpoint.
Methods
Five established methods
Each method follows its own epistemology. QualiLens implements distinct pipelines with method-appropriate stages, checkpoints, and figures.
Grounded Theory
- Stages
- Familiarization, open coding, axial coding, selective coding
- Checkpoints
- Code review, core category, theoretical model
- Report figure
- Paradigm flow model
Thematic Analysis
- Stages
- Familiarization, initial coding, theme construction, theme review
- Checkpoints
- Code review, theme refinement
- Report figure
- Thematic map
Content Analysis
- Stages
- Codebook derivation or import, systematic coding
- Checkpoints
- Codebook approval, frequency review
- Report figure
- Frequency bar chart
Framework Coding
- Stages
- Apply framework, chart by source, build matrix
- Checkpoints
- Low-confidence review, emergent codes, matrix
- Report figure
- Framework heatmap
Literature Synthesis
- Stages
- Structured extraction, cross-paper synthesis, concept matrix
- Checkpoints
- Extraction table, concept review
- Report figure
- Concept-by-paper heatmap
Built for researchers
Each method follows published methodological commitments. The wizard records your choices verbatim, so you have a record for the methods section.
Explore all featuresWhy QualiLens
What makes QualiLens different
Established CAQDAS platforms host your data on their servers or require per-seat licenses. QualiLens is open source, runs locally, and replaces manual coding with auditable LLM-assisted coding behind researcher-controlled checkpoints.
Researcher-led checkpoints
The pipeline pauses at each analytic decision point. Your edits are final and are not overwritten by later automated stages.
Provenance on all excerpts
Each coded passage stores the verbatim quote and its character position. The coded-source reader draws coding over each document in place.
Full audit trail
All stage boundaries, model calls, and researcher decisions are logged. The report appendix reproduces this trail for reviewers.
Runs locally
No QualiLens server. Your data go only to the AI provider you chose, over a direct API connection. Nothing is collected.
Four providers, your key
Anthropic, OpenAI, Google, and Mistral. Cost estimation before each run, and resumable runs that don't re-bill finished work.
Reports and Word export
Interactive report with narrative sections, method-appropriate figures, full evidence chains, and one-click Word export.
Your data stay on your computer.
No QualiLens server, no telemetry, no analytics, no background connections. Analysis calls go directly to the provider whose key you supplied — and nowhere else.
Read the full privacy statementReady to start?
QualiLens requires a Mac with Apple silicon or a Linux computer (on Windows, WSL), Python 3.11+, and an API key from any supported provider. A single-document analysis typically costs a few cents.