Coding you can hear, speak, and drive without a mouse.
Accessible Claude gives you the full power of Claude Code, rebuilt for developers who are partially sighted, low-vision, or legally blind. In 1.0 it crosses the real frontier: the AI does the looking — reviewing the diff, touring the repo, reading the failing build — and narrates it back, while you drive the whole app by keyboard or voice.
Free · bring your own key · Windows · NVDA & Narrator verified
The AI does the looking, so you don’t have to.
A screen reader that reads, keyboard nav, high-contrast themes — that’s table stakes, and Accessible Claude has always had it. 1.0 crosses the real frontier: using the AI as your eyes and your hands. It does the seeing — reviewing a diff, orienting in a new repo, reading a failing build — and narrates it back to you.
Change Review
Walk me through what you just changed.
The glance you’re missing
You can’t scan a diff. When an agent edits ten files, verifying its work non-visually is the single hardest, most trust-critical moment in the loop.
A navigable, spoken review of the agent’s own edits — drill from turn to file to change. It narrates the structure (“three files changed, 40 lines added”), the content, and an on-demand “why” for any edit. Undo at a coarse grain if you don’t like what you hear. Sourced from the agent’s edits directly — no git required.
Codebase Orientation Tour
Give me a tour of this project.
The glance you’re missing
Dropped into an unfamiliar repo, a sighted developer skims the file tree and gets their bearings in seconds. Without that glance, orientation is slow and disorienting.
A narrated, queryable map of the codebase — architecture, entry points, the key modules, how data flows. Navigable and conversational: “where is authentication handled?”, “what calls this function?” You build a mental model by asking and listening, not by tabbing blindly through hundreds of files.
Narrated Run & Verify
Run it and tell me what happened.
The glance you’re missing
“Did it pass?” shouldn’t mean paging a screen reader through a thousand lines of build and test log to find the one line that matters.
The agent runs your build, tests, or app and narrates the outcome — passed or failed, the single error that actually matters, and a suggested fix — instead of dumping logs. The raw output is always one command away when you want it. You get the signal, not the noise.
Eyes-free. Now hands-free too.
Hold one command key and drive the entire app by speech — navigate, move through lists, control reading, and act. It’s a deterministic, offline grammar, not a chatbot: instant, private, and it echoes anything risky back for a one-word confirmation before it happens.
How voice commands work- navigatego to settingsinstant
- move & readnext changeinstant
- actapproveconfirmed
A second pair of eyes, only when you want them.
Invite a sighted teammate into your live session and decide exactly what they can do — watch, send prompts, or approve actions — as separate permissions you can take back instantly. It’s end-to-end encrypted, every action is announced to you, and if the link drops, control returns to you alone.
How pairing works- viewwatches your session, read-onlyrevocable
- sendcan propose prompts you runrevocable
- approvecan approve or deny tool actionsrevocable
Everything Claude Code does — accessible by design.
Under the 1.0 headline features sits a mature accessible foundation: the interface, the pairing model, the voice layer, and the details that keep it yours.
Built for the way you work
Learn it without sight, discover commands without scanning a terminal, and keep your keys where they belong.
- A guided tutorial that’s authored and fully offline — no key, no network — offered on first run and resuming where you left off
- Inline “/” slash-command autocomplete in the prompt box — find commands like /speckit-plan without memorising them
- Bring your own key, stored in the operating system’s secure storage — never in plain files, the transcript, or speech
Learn it in minutes — by doing.
A hands-on, screen-reader-friendly walkthrough teaches the layout and landmarks, running the agent, approvals, voice, and the three “agent as your eyes” lenses. It’s authored and fully offline — no API key, no network — offered on first run, and it resumes exactly where you left off. Nothing is gated behind sight.
- Offered on first run
- No key, no network
- Resumes where you left off
Frequently asked
Who is Accessible Claude for?
Developers who are partially sighted, low-vision, or legally blind — and the sighted teammates who pair with them. It’s built around large, high-contrast type, the keyboard, and the screen reader, rather than treating them as an afterthought.
What does “the agent as your eyes” mean?
It’s the headline of 1.0: three lenses that do the visual work non-visually. Change Review gives a navigable spoken walkthrough of the agent’s own edits; the Codebase Orientation Tour narrates and answers questions about an unfamiliar repo; and Narrated Run & Verify runs your build or tests and tells you the one error that matters instead of dumping logs.
Can I really drive the whole app by voice?
Yes. New in 1.0, you hold one command key and speak to navigate, move through lists, control reading, and act — a deterministic, offline grammar rather than a chatbot. Navigation and reading fire instantly; anything risky, like approving a shell command, echoes back exactly what it will do and waits for a one-word confirmation, and can never be triggered by a misheard word.
How is it different from Claude Code?
It runs the same agentic coding sessions with the same commands and behaviour as Claude Code, but as a desktop app rebuilt for accessibility: fixed high-contrast themes, an ultra-large font, full keyboard operation, verification against NVDA and Narrator, and built-in dictation and read-aloud.
Does it work with my screen reader?
Yes. Accessible Claude is verified against NVDA and Narrator on Windows, and everything is operable by keyboard alone. Its own read-aloud coordinates with the screen reader you already use rather than fighting it.
What does it cost?
It’s free, and you bring your own key — your own Anthropic credentials for the model, and optionally your own OpenAI or ElevenLabs key for voice. Keys are stored in the operating system’s secure storage.
Can a sighted colleague help me?
Yes. You can invite a sighted teammate to your live session and choose exactly what they can do — view, send prompts, or approve tool actions — as separate permissions you can revoke at any moment. Pairing is end-to-end encrypted and you always stay in control.
Which platforms are supported?
Windows today. Join the waitlist and we’ll let you know as support for other platforms lands.
Code the way that works for you.
Free, and it uses your own API key. Download for Windows, or join the waitlist for updates and other platforms.