eri

An agentic editor that builds at the speed you think.

eri listens to your voice, watches where you point, and remembers how you build. It gets sharper every session you give it.

Download for macOSRequest access
Chapter 1

An IDE that watches, listens, ships.

eri · pricing-app

⌘K to ship
1// app/pricing/page.tsx
2export default function Pricing() {
3 return (
4 <section className="pricing">
5 <h1 className="text-4xl">Pricing that scales</h1>
6 <Toggle value={annual} />
7 <Price value={annual ? 240 : 25} />
8 <Button>Start free trial</Button>
9 </section>
10 );
11}

""

Preview · live

looking · h1.hero-headline

Pricing that scales with you

One plan. Predictable cost. Cancel any time.

MonthlyAnnual
$25/moper workspace

No card. 14 days. Local-first.

main·page.tsx·line 1, col 1·TypeScript React
eri engine · diffusion
Chapter 2

The loop. Look. Speak. Ship.

The eri loop

Your eyes stay on the prototype.
The code rewrites itself.

Look at the part you want to change. Say it out loud. The diffusion engine streams the patch into the file and the prototype updates under your gaze. Voice, vision, completion, and apply all run on your hardware — the planner is the only call that crosses the wire.

Prototype · live preview

looking · h1.hero-headline

Pricing that scales with you

One plan. Predictable cost. Cancel any time.

MonthlyAnnual
$25/moper workspace

No card. 14 days. Local-first.

pricing.tsx — eri

eri engine
1<section className="pricing">
2 <h1 className="text-4xl">Pricing that scales</h1>
3 <Toggle value={annual} />
4 <Price value={annual ? 240 : 25} />
5 <Button>Start free trial</Button>
6</section>

""

gaze → voice → diffuse → render

Only in eri

The reason to switch

The four primitives that turn the loop on the left into a single gesture — and the ones nobody else has shipped.

  • Voice on existing repos

    Whisper-on-device, sub-second

  • Gaze targeting

    MediaPipe in-process, camera frames stay local

  • Diffusion edits

    The patch resolves; no left-to-right wait

  • Local Tab

    Embedded inference on your hardware

Tab, ⌘K, Composer, agents, @-mentions, MCP — all in the box. The four above are the part nobody else ships.

Where the work happens

One price. Hard cap. No credit pools, no auto-routing surprises, no metered overage.

three columns · one boundary

Local-first

The editor does the work on your hardware.

Whisper transcribes voice. MediaPipe reads your gaze. An embedded inference runtime drives Tab and fast-apply. Your semantic index lives on your disk. Nothing in this list ever leaves the machine.

  • Voice transcriptionWhisper, in-process
  • Gaze trackingMediaPipe, on-device
  • Tab completionEmbedded llama.cpp
  • Apply / diffLocal fast-apply
  • Semantic index~/.eri/index.sqlite
  • TTS playbackCached presets, no upload

Proof

~/.eri/index.sqlite

Bundled, not metered

Every feature is in the price. There is no second meter.

Tab, diffusion edits, semantic search, voice, gaze, MCP, checkpoints — all included. Tab runs in-process on your machine, so the heaviest call in the loop has no per-token cost. Subscribe once; nothing else opens a tab in your billing console.

  • Tab completion — Local engine, bundled in subscription
  • Diffusion edits — Multi-line patches, one resolve
  • Semantic search — Index lives on your disk
  • Voice + gaze — Always available, no add-on
  • MCP servers — Plug in tools you already trust
  • Checkpoints — Rewind any agent turn locally

Proof

Tab runs via embedded inference. No per-token cost to you.

Honest about cloud

One thing crosses the wire. We tell you which one and when.

Planning and multi-file reasoning route to the diffusion engine — and only that. Every call shows the model, the token count, and the wall-clock latency in the iterator panel. Flip on audit and every call lands in a log file you own.

  • Crosses the wirePlanner only
  • Visible per turnModel · tokens · latency
  • Off by defaultTelemetry
  • Opt-inAudit log

Proof

~/.eri/audit.log

Every shortcut you have muscle memory for.

Plus the muscle eri brings on its own — predicting your next move, learning your patterns, remembering you across sessions. Your settings.json, keybindings, and extensions carry over on first launch.

See the full feature map
Chapter 3

Three lazy words. One precise prompt.

Context pipeline

Three lazy words. One precise prompt.

You speak in fragments. eri reads your gaze, your last few turns, and the file you're actually in — then compiles a prompt the model can act on, and routes it to whichever brain handles that kind of work best.

01Input— signals eri already has
"make it glow"
FREE
3 words · captured in 410ms
Toggle.billing-cycle
FREE
pricing.tsx · line 47
recent turns
STARTER
added annual toggle 2 turns ago · 4 props on draft
codebase context
PRO
ToggleProps · tokens.accent.neon · sibling: PriceCard
02Compile— assemble + ground
assembling
0%
compiled prompt

▌

voice intentFREE
gaze anchorFREE
tokens lookupPRO
tone presetPRO+
03Route— pick the right brain
eri router
crates/eri_router
eri's engineFREE
diffusion · short edits
Claude Sonnet 4.6STARTER
polish · careful diffs
Claude Opus 4.7PRO
reasoning · long horizon
Gemini 2.5 ProPRO+
multimodal · image-in
decision
waiting on compiled prompt…
04Preview— what you're looking at
pricing.tsx · <Toggle>
live preview
Billing cycle
Monthly
Annual
$19/mo · billed annually
05What changed— diff applied
awaiting model
—
+ shadow: 0 0 18px var(--accent);
~ Toggle.tsx · 1 line modified
effect
diffusion denoising · preview will resolve on lock.
FREE baseline pipelineSTARTER iteration historyPRO codebase brain · auto-routePRO+ multimodal · voice override

Build at the speed you think.

The agentic editor for engineers who think faster than they type. Bring your own keys. Local-first.

Download for macOSRequest access→