Read the PDF. Go deeper without leaving it.

Turn passages into
maps of understanding.

Lensmap is not a full-document AI autopilot. Start from the passage you are actually reading, expand only the context you need, and keep what you learn as Maps that lead back to the source.

Keep using Chrome’s built-in PDF Viewer.

PMBOK analyzed with LensmapSelect a passage → Ask → Keep the Map
01 · HIGHLIGHT · PDF p.29Selected

PMBOK 7 · The tailoring process

The tailoring process

Projects in one organization usually have a lot in common.

First, a high-level partial tailoring for all projects. Then, a final tailoring for each project.

The shared structure reduces duplicated effort while leaving room for each project.

02 · ASK

What does this two-stage tailoring look like as a practical flow?

S1 · PDF p.29Expand only what is needed
PMBOK · MethodologiesComparison
PMBOK · 80/20 ruleChart
03 · MAP · Tailoringselection + needed references

All examples come from the same public PMBOK PDF. The Map changes shape—definition, table, comparison, flow, and more—based on the question.

See the PMBOK case studies →

Focus → Expand → Map

You stay in control of the reading.

  1. 01

    Focus

    You choose what deserves attention. Start from a passage or visual region instead of asking AI to read the whole document for you.

  2. 02

    Expand

    When the selection is not enough, Lensmap can read nearby blocks, the same section, or related passages—and records what it added.

  3. 03

    Map

    What you learn is retained automatically as a Map: diagrams, tables, comparisons, concise explanations, and references in the form that fits the idea.

Chat is the scratchpad. The Map is the outcome.

Come back later and still know why it made sense.

The point is not to make chat logs look prettier. A Map keeps how reading became understanding. You can expand only the context needed from one passage or synthesize a whole book into a multi-block understanding Map. The examples below use a real public CC BY project-management PDF.

PMBOK ANALYZED WITH LENSMAP

Tailoring moves from organization-wide to project-specific.

Start from the PMBOK passage that recommends high-level tailoring across the organization followed by final tailoring for each project. Ask for a practical sequence, and the related context becomes a five-step flow.

1
passage selected by the reader
+1
additional context read by AI
2
References retained in the Map
A five-step flow from organization-wide tailoring to project-specific adjustment
The starting passage and the related context read for the answer remain available as source references.Swipe to inspect the actual Map →

SAME PMBOK, DIFFERENT QUESTION

Change the question, and the Map changes shape.

Selecting four passages on P3.express, PRINCE2, Scrum, and PM² and asking how the book positions them produced a four-column Comparison Map. Lensmap does not force every insight into one diagram type.

Comparison of P3.express, PRINCE2, Scrum, and PM²
Another real Map: four referenced passages are organized into a four-way comparison.
ExamplesNine Maps from focused passages, plus one PMBOK 7 Insight Map.View all →

Example source: Nader K. Rad, PMBOK® Guide 7, Underneath the Surface — CC BY. The Map is a Lensmap-generated summary and structure. The book reflects the author’s personal views, not PMI’s. PMBOK® is a registered mark of Project Management Institute, Inc.

Keep the references

Lensmap distinguishes what you selected from context added by AI. Citations lead back to the original PDF, so the explanation never becomes detached from its source.

Local-first

PDFs, indexes, Explore threads, and Maps stay on your Mac. Lensmap has no Lensmap-owned cloud or relay server; only the references needed for an answer are sent to AI.

Local-first · ChatGPT subscription

No API key. Use Codex within your ChatGPT subscription.

Lensmap has no Lensmap-owned external server. PDFs, indexes, and Maps stay on your Mac, while the local Lensmap Server connects to Codex App Server. Only the references needed for AI processing are sent to OpenAI through Codex. No separate API key is required, and usage stays within the Codex allowance included in your ChatGPT plan.

YOUR MAC · CHROMELensmap Side PanelFocus / Explore / Maps
YOUR MAC · LOCALLensmap ServerPDF index · references · retrieval
YOUR MAC · CODEXCodex App ServerAuthenticated Codex session
  • No Lensmap-owned cloud or relay server; PDFs, indexes, and Maps stay local
  • Only the references and constrained context needed for the answer go to AI
  • No API key; use Codex authenticated with your ChatGPT account

Open source · Apache-2.0

Try the current build from GitHub.

Development setup lives in the README. This page keeps the installation section to a single entry point instead of repeating the steps.

FAQ

Frequently asked questions

Which PDF reader does Lensmap use?

Chrome’s built-in PDF Viewer. You keep the PDF in Chrome and use Lensmap in the Side Panel beside it.

Does Lensmap upload the entire PDF to AI?

No. PDFs and indexes stay on your Mac. Codex receives the references needed for the current question plus constrained additional context only when necessary.

Is my PDF sent to a Lensmap server?

Lensmap has no Lensmap-owned external server or cloud. References needed for AI processing go from the local Lensmap Server through Codex App Server to OpenAI.

Which AI does it use?

Lensmap uses Codex authenticated through ChatGPT. No separate API key is required; usage is within the Codex allowance included in your ChatGPT plan.

Can it read local PDFs?

Yes. Enable “Allow access to file URLs” in the Lensmap extension details in Chrome.

What platforms are supported?

The current packaged target is macOS Apple Silicon with Google Chrome 141 or later.