Read Skim Aloud — TTS Built for Academic PDFs

Skim is the open-source PDF reader that grad students, researchers, and PhDs swear by — highlighter-first, keyboard-driven, no distractions. CastReader adds the one thing missing: natural voice TTS with word-level highlight on the real Skim window. Hear your papers while your eyes scan figures. Native Accessibility path — no Screen Recording.

macOS native · Free · No login · 40+ languages · Native AX, no OCR

paper.pdf — Skim

We introduce a sparse mixture-of-experts architecture that scales parameter count decoupled from inference cost. Empirically, our 7B-active / 70B-total model matches dense 30B baselines on reasoning benchmarks while reducing per-token FLOPs by 4.1×. Ablations isolate the contribution of load-balancing loss versus auxiliary router regularization.

Highlight follows each spoken word directly inside your Skim window.

Reading selection…1.0×

One App Covers Every Mac PDF & Research Reader

Same gesture, same word-level highlight — across every native PDF app researchers and students actually use.

Preview

Apple's PDF viewer

PDFgear

Free PDFKit reader

Adobe Acrobat Reader

Industry standard

Skim

Academic favorite

PDF Expert

Readdle paid PDF

Three Steps — Native, No OCR

Skim is an open-source Cocoa app built on Apple's PDFKit and exposes text through macOS Accessibility. CastReader reads it directly — no screen capture, no OCR, pixel-perfect highlight.

1

Install

Download CastReader for Mac (.dmg). Grant Accessibility permission once. Skim doesn't need Screen Recording permission.

2

Select

Open any paper in Skim. Drag-select the section you want to hear — abstract, method, related work, discussion, conclusion.

3

Click ▶

A play button appears next to your selection. One click — audio starts and the highlight glides inside the actual Skim window, word by word. Skim's own highlighter keeps working.

When Academics Add Read-Aloud to Skim

Skim is where the heavy academic reading happens — papers, theses, chapters, preprints. A listening channel cuts session fatigue in half.

Triage paper abstracts at 1.5×

Screening 10 papers a day for your literature review? Listen at 1.5× while eyes scan the figures — triage 5× faster without losing signal.

Absorb method and results sections

The middle of a paper is where comprehension matters most. Audio at 1× forces engagement; lets you actually think about experimental design instead of word-decoding.

Listen + Skim's highlighter = ultimate study loop

Listen to a paragraph. If something clicks, grab Skim's highlighter. CastReader doesn't touch Skim's annotation layer — both tools stack.

Proof-listen your own thesis draft

Export your thesis chapter to PDF, open in Skim, listen. Hearing your prose catches awkward transitions and dangling modifiers that eyes skim past.

Read non-English academic PDFs

Chinese technical whitepapers, Japanese lab reports, German mathematical texts — CastReader auto-detects and switches to native voices at sentence boundaries.

Accessibility for dyslexia and long-session fatigue

Six hours of dense academic PDF reading strains anyone. For dyslexic researchers, audio + highlight doubles comprehension speed.

Why CastReader Fits Skim's FOSS Academic Ethos

Skim is free and open source because academics believe tools for thought shouldn't charge rent. CastReader shares that principle.

Pixel-perfect highlight on the real document

Transparent overlay tracks each spoken word on the actual Skim PDF — not a clone pane or sidebar. Pixel-aligned via AX.

40+ natural voices, CJK and multilingual

Kokoro-quality voices across English, Chinese, Japanese, Spanish, French, German, Korean — far better than macOS legacy voices.

Variable speed, click to jump

1.5× for lit-review triage, 0.9× for careful method reading. Click any word to resume from there — no re-listening to paragraphs.

Free forever — no academic subscription trap

No account, no seat license, no daily word limit. The same free-forever principle as Skim itself.

Common Questions

How do I read a PDF aloud in Skim?

Install CastReader for Mac, grant Accessibility permission, open any PDF in Skim, drag-select a paragraph, then click the floating ▶. Word-level highlight appears on your real Skim window.

Does Skim have built-in text-to-speech?

Skim doesn't ship its own TTS. You can use macOS Speak Selection (legacy voices, no highlight, no resume). CastReader adds Kokoro-quality voices, word-level highlight on the real Skim window, and variable speed — everything the built-in lacks.

Will CastReader conflict with Skim's highlighter?

No. CastReader only triggers on explicit text selection + ▶ click. Skim's highlighter, notes, snapshots, and reading-mode all work untouched.

Does it work on arXiv / bioRxiv / PubMed papers?

Yes. Once you download the PDF and open it in Skim, any selection reads aloud. For reading those sites on the web, see our dedicated pages for arXiv, bioRxiv, and PubMed.

What about multi-column academic layouts?

Skim's column-aware selection works well — drag-select the column you care about and CastReader reads in order. For very dense two-column layouts, select one column at a time for clean reading flow.

Does it handle LaTeX or math equations?

CastReader reads the literal characters of your selection. Inline LaTeX or plain-text math is spoken as characters. For rendered equation images, select the surrounding prose instead — the equation is usually explained in that paragraph.

Can I listen to Chinese academic PDFs?

Yes. CastReader auto-detects CJK and switches to native Mandarin / Japanese / Korean voices. Mixed-language papers (English abstract + Chinese discussion) switch voice at sentence boundaries.

Is it really free?

Yes. 100% free — no account, no credit card, no premium voice gate, no daily word limit. Same FOSS spirit as Skim.

Ready to Hear Your Papers Speak?

Install once. Open any paper in Skim. Select. Click ▶. Listen. Free for every grad student, every researcher, forever.