Skip to main content
Vinci by SimpleDirect®

AI for a life beyond the screen.

Draft a follow-up, compare options, or turn messy notes into a plan.

Try an example

Demo

Prewritten examples

Available now

You: Draft a warm follow-up after a client meeting

VinciMezzo

Hi Sam, Thanks for meeting today. I’ll send the revised schedule by Thursday. Could we use our next call to confirm the launch date and who owns each step? Maya

Try a follow-up

Try in Vinci

Type a question to open it as a draft in Vinci.

Pay for what you use. Each charged reply shows its cost.

Provider cost + 15%

Watch Vinci Code CLI complete a task
Try it

See what Vinci does, not just what we say.

Four short, scripted demos you can run right here. Each one says whether the feature is available now or still coming.

DemoAvailable now

Research with sources

Ask a question and get an answer with numbered citations you can check.

You: Do heat pumps work in a cold climate?

Forte
Yes, with the right model. A heat pump moves heat from the outdoor air into your home instead of burning fuel to create it 1. Cold-climate models keep heating at low outdoor temperatures, though their efficiency drops as it gets colder 2. Many homes in cold regions keep a backup heat source for the coldest days 13.

Sources

  1. 1How air-source heat pumps worknorthwind-research.example
  2. 2Heat pumps in cold climatesharbourline-energy.example
  3. 3Backup heat for the coldest daysmaplecrest-homes.example

Demo sources (fictional)

DemoComing soon

Hand a task to Vero

In this scripted demo, Vero explains each step in plain words and stops when something needs your yes.

Suggest a time for Northwind Studio's Q4 review and draft the invite.

  1. Vero reads the request and repeats it back so you can correct it.
  2. Vero works through the request and finds two possible times.
  3. Vero compares them and picks Thursday at 10:00.
  4. Sending the invite would email three people, so the demo stops here and asks you first.

Preview: this is where Vero stops and waits for your yes. Nothing is sent until you approve. Hand it to Vero to watch each step.

DemoComing soon

Coding tasks in Vinci for Work

A preview of a task that fixes a failing test: brief, steps, diff and test result.

Task preview · northwind-portal (fictional)

Fix the failing sort test

Done

Replay · no code runs

  1. Brief

    Fix the failing sort test in northwind-portal without changing the public API.

  2. Steps

    1. 1Read src/sort.ts and test/sort.test.ts
    2. 2Ran the sort test: 1 failing, expected [3, 2, 1], received [1, 2, 3]
    3. 3Changed only the comparison in src/sort.ts
    4. 4Ran the sort test again, then the full suite
  3. Result

    src/sort.ts

    • - return items.sort((a, b) => a - b);
    • + return items.sort((a, b) => b - a);

    24 tests passed · 0 failed

    One line changed in src/sort.ts. The focused test and the full suite pass.

DemoComing soon

Vinci Apps

Small apps Vinci can build and run for you. Try these early versions right here.

2 left

Long-running workComing soon

Hand off long-running work. Stay in control.

A preview of the project workspace we are building, from our design mock-ups with a fictional company. The demo shows a run that stops for your approval and finished work that shows its sources.

Preview Vinci for Work

Fictional company and data. Built from design mock-ups; the shipped product may differ.

Walk through the projectDemo

Project overview · Northwind Studio

Q4 Client Review

Quarterly portfolio review for 14 retainer clients: health, renewals and a board-ready summary by Dec 12.

Live runs
3
working now
Waiting on you
2
approvals before sending
Artifacts
9
3 new this week
RunStatusProgress
Compile Q4 portfolio review for 14 clientsReport and 12-slide board deckRunningStep 7 of 11
Draft renewal emails for at-risk accounts3 drafts ready to sendNeeds approvalPaused at step 9
Weekly client health digestMondays at 8:00Completed11 steps · 14 min
Summarize six client kickoff callsNotes and action itemsCompleted6 steps · 9 min

Run timeline

Compile Q4 portfolio review for 14 clients

Step 7 of 11
  1. Read the brief and 14 client folders
  2. Pulled Q3 and Q4 revenue for 14 accounts
  3. Scored client health from time tracking and invoices
  4. Drafted the report outline and built 4 charts
  5. Writing “Renewals and risks”Live

Vero is narrating

Atlas Freight and Kettle & Co. dropped below 60% utilization in October, so I'm marking them at risk and citing the October timesheets for each number. Next I'll draft a short renewal note for each at-risk account.

Needs your approval

Send 3 emails as Maya Chen?

In this demo, sending email waits for your approval.

  • Atlas FreightRenewal check-in before Dec 31
  • Kettle & Co.Q1 scope and retainer options
  • Bramble HealthQuick sync on the October report

Steer

  1. Lead the board deck with renewals and keep it to 12 slides.

  2. Got it. Renewals open the deck, and I'll fit the rest into 12 slides. Steps 9 and 10 now follow that order.

Verified artifact

Q4 Portfolio ReviewReport · v3

Run finished · checks passed

All 42 figures match their sources. I left 2 notes where the CRM and timesheets disagreed, so you can decide.

Verification

  • 42 of 42 figures traced to a source
  • Totals reconcile with invoices
  • Charts regenerated from data
  • 2 notes for your review
ClientUtilizationHealth
Lumen Dental91%Healthy
Bramble Health86%Healthy
Orchard Foods79%Watch
Atlas Freight54%At risk
Kettle & Co.57%At risk

Sources

  1. 114 client foldersProject files
  2. 2Deals export, Q3 to Q4CRM
  3. 3October timesheetsWorkspace
  4. 4Rate cards 2027Project files
Coming soonDemo

Follow a project from your phone

A mock-up of following a project from your phone.

The Vinci family

One Vinci, from your browser to your terminal.

Start where you are. Each card says what you can use today and what is still coming.

Vinci app icon
Available now

Vinci in your browser

Draft a message, explain an error or turn notes into a plan, with Mezzo for everyday tasks or Forte for deeper work.

Open Vinci
Available nowPublic beta

Vinci Code

Vinci Code is a coding agent in your terminal. It reads your project, edits files and runs commands, and asks before destructive ones.

See Vinci Code
Coming soon

Vinci for teams

See the whole team's runs and approvals in one place, with roles for each person.

Team runs in progress

Demo
  • Compile Q4 portfolio reviewMaya Chen · Running · 7/11
  • Shortlist 40 portfolio applicantsJonah Okafor · Running · 4/9
  • Collect forms and set up 6 new vendorsSam Lefebvre · Needs approval
Preview Vinci for Work
Pricing that stays on your side

Provider cost, plus 15%. That’s it.

Mezzo and Forte are Vinci’s two live model classes. We show what inference costs us, what you pay, and keep the incentive simple: when our cost goes down, yours does too.

Fast, everyday work

Vinci Mezzo

Runs GLM-5.3-Flash today

The default for chat, writing, planning and lighter research. Move to Forte for code and file work.

Input

Output

DeepInfra rate

$0.15

$0.50

You pay with Vinci (+15%)

$0.1725

$0.575

USD per 1 million tokens. The model and provider behind each Vinci class are published and may improve over time.

Code and deeper work

Vinci Forte

Runs GLM-5.3 today

For coding, files and tasks that benefit from a stronger reasoning model.

Input

Output

DeepInfra rate

$0.90

$4.00

You pay with Vinci (+15%)

$1.035

$4.60

USD per 1 million tokens. The model and provider behind each Vinci class are published and may improve over time.

The Costco model for AI

A small, fixed markup. Better value over time.

Vinci adds 15% to inference cost—not a mystery multiplier. That keeps us working on the same side of the bill as you.

Lower cost, passed through

We keep looking for lower-cost inference from vetted providers.

Better models, same classes

We regularly review and upgrade the models behind Mezzo and Forte as quality improves.

The same models we use

The Vinci team uses Mezzo and Forte every day to chat, code and work.

Want the broader cost comparison?

Open the monthly examples for casual, developer and power-user workloads.

Compare Vinci with

Casual user

Everyday help

1M input + 250K output tokens each month

Vinci Mezzo

$0.3163

GPT-6.1 Terra

$2.25

Claude Sonnet 5.5

$4.50

86% lower reference cost than the least expensive model shown.

Developer

Regular coding work

10M input + 2M output tokens each month

Vinci Forte

$19.55

GPT-6.1 Sol

$40.00

Claude Sonnet 5.5

$40.00

51% lower reference cost than the least expensive model shown.

Power user

High-volume work

50M input + 10M output tokens each month

Vinci Forte

$97.75

GPT-6.1 Sol

$200.00

Claude Sonnet 5.5

$200.00

51% lower reference cost than the least expensive model shown.

Mezzo and Forte use DeepInfra’s published standard rates plus exactly 15%. Vinci charges the provider-reported inference cost plus 15%, so temporary provider discounts can make an actual request cost less.

The same token volume is not the same capability. Models, tokenizers, tools and results differ; test your own work. Caching, tools, taxes, long-context and regional charges are excluded.

One messageDemo

What one short reply costs.

A short Mezzo reply at the reference rates above. In Vinci, each charged reply shows its cost.

Transparent task receipt

Vinci Mezzo · GLM-5.3-Flash

1 call · 2,000 in / 4,400 out · 0% cached

New input

2,000

$0.15 / 1M

$0.0003

Cached input

0

$0.03 / 1M

$0.0000

Output

4,400

$0.50 / 1M

$0.0022

Provider subtotal$0.0025

Vinci · 15%$0.000375

You pay$0.002875

Illustrative usage for this browser demo. No provider request was made; real receipts use the tokens and costs reported for the actual task.

Vero, the Vinci character, waving hello
Meet VeroComing soon

Meet Vero.

Vero is the character who will explain what Vinci is doing while work runs: what it is reading, what it decided and when it needs you.

  • Says what is happening in plain words
  • Stops and asks before anything consequential
  • Shows the finished work with its sources

Vero is coming to Vinci.

Read about Vero

A small first step

Start with one useful task.

You do not have to move your whole workflow to see whether Vinci helps.