A research workspace
you can actually steer.
CodexIQ captures what you read, connects it as your knowledge grows, and synthesizes it through Prisms — weighted lenses that tell the AI what to emphasize. Your model. Your keys. Your data.
Most AI gives you the average answer. Prisms give you yours.
A Prism is a weighted lens. Attach one to a topic — or just click the words that matter inside a note — and CodexIQ bends every summary, connection, and insight toward what you've emphasized, and away from what you haven't.
- Click to weight. Emphasize or de-emphasize any phrase. Thickness shows weight; color shows topic.
- It consolidates for you. Related weights get suggested as a single named Prism you can reuse.
- Nothing is silent. Every output shows exactly what shaped it — and lets you tune it on the spot.
Capture, connect, and think — without the busywork.
A structure that mirrors how ideas actually grow, with AI that works in the background instead of waiting for prompts.
Glyphs, Nodes & Codexes
Notes become Glyphs, Glyphs cluster into Nodes, Nodes live in Codexes. Structure that grows with your thinking instead of fighting it.
Insights & Echoes
CodexIQ surfaces patterns you didn't ask for — and quietly links new material to the threads you already care about.
The Lens
A floating companion that's always a click away — capture what you're reading, explore a tangent, or ask for help, from anywhere.
Intent & Pursuits
Point the system at a question — for one session with Intent, or for months with a Pursuit that tracks your progress toward an answer.
Ingest files & audio
Drop in PDFs, Word docs, and recordings. Text is extracted, speech is transcribed — and the original binary is discarded.
Session Brief
Come back to a calm digest of what changed while you were away — ranked by the Prisms and Watch List you've told it matter.
Four moves, then it compounds.
Capture
Read anything. Paste text, drop a URL, or add a file as a Glyph — clean text, source kept, nothing lost.
Connect
CodexIQ links Glyphs into Nodes and Codexes and surfaces Echoes between things you'd never have filed together.
Weight
Set Prisms and Intent to tell the AI what matters — by topic, by phrase, or by the question you're chasing.
Synthesize
Get summaries and Insights refracted through your weighting — with a receipt of exactly what shaped each one.
Calm by default. Dark when you want it.
A quiet, keyboard-first interface that gets out of the way. The same workspace, the same Prisms, the same receipts — in a theme that's easy on long sessions.
- Smart Panel keeps connections, Prisms, and fresh insights one glance away.
- Find anything across every Codex with semantic and exact search.
- Your call on AI — a usage meter, web search toggle, and local-only logic switch, always in reach.
Private by architecture, not by promise.
CodexIQ is built so your material — and your model access — never become someone else's product.
Bring your own key Coming soon
Use your own OpenAI, Anthropic, or Gemini account — or run entirely local with Ollama. Keys are AES-encrypted at rest, never logged, never returned in exports.
Text in, files gone
Uploads are extracted to text and metadata, then the original PDF, doc, or recording is discarded. CodexIQ never stores your binaries.
Nothing trains anyone
Your Glyphs, Prisms, and Insights are yours alone. No shared training, no cross-account mining — a local-only logic mode is one click away.
A note app remembers. A chatbot answers. CodexIQ does both — on your terms.
Start your first Codex today.
Capture what you're reading, weight what matters, and let the synthesis come to you.