Private model access

State-of-the-art models. None of your data leaves with them.

Kluff gives your team leading models without sending workspace data to OpenAI, Anthropic or Google.

  • No training on your data
  • PII redacted first
  • Regional processing

From capability to everyday work

Private model access, in practice.

Kluff separates the team’s workspace context from the model choice. Model access and data-handling controls are part of the platform, while projects keep the context and working results together.

Use it for
Teams evaluating where AI processing happens and how sensitive context is handled.
Start with
Review your intended data types, deployment requirements, and model-access configuration with the Kluff team.
Read the setup guide
Illustrative workflow Sample data
The request

“Review how a customer-support request reaches a model.”

  1. Start with the information available to the person asking.

  2. Apply configured data-handling controls before model processing.

  3. Keep the answer beside the original work in the project.

Data-handling review

Example output
Context
Which information enters the request?
Processing
Which model path is configured?
Review
What redaction and access controls apply?

Guarantees

Rent the intelligence. Keep the data.

Stays in your workspace

Prompts and context are not sent to the big model vendors.

No lock-in

Your memory lives in Kluff. Models can change underneath without losing context.

Easy security reviews

One clear answer for buyers and reviewers on where data goes.

How it works

What happens to a prompt

1

Built in your workspace

Kluff assembles the prompt from the thread, memory and tools.

2

Redacted

Personal data is removed by a local model.

3

Answered privately

The request runs on Kluff's private model access.

4

Recorded

The result lands in the thread with an artifact.

FAQ

Questions, answered.

Evaluate private model access with the setup, access, and review points in mind.

Teams evaluating where AI processing happens and how sensitive context is handled.

Review your intended data types, deployment requirements, and model-access configuration with the Kluff team.

Confirm the model path, redaction settings, and data-handling requirements for your workspace before adding sensitive material.

Start with one project and a clear task. Add the files or authorized accounts the task needs, then review an initial run before scheduling recurring work.

Integration actions use your owned or shared connections. Safe mode pauses actions that have not been pre-approved. MCP connections, canvas apps, and Computer Use have their own access controls.

No. Kluff uses private model access, and PII is redacted before any model call.

Business starts at $15 per user per month on monthly billing for teams of up to 25, with a 14-day free trial. See Pricing for other team sizes, annual billing, fair use, and Enterprise options.

Unlimited inference.
Company memory.

Connect your tools, ask your first question in a project, and watch memory, spend, and visibility stay in one place—not scattered across personal chats.

$15 / user · Unlimited AI
Q4 launch› Launch dashboard
Sara Castillo

@kluff build a launch dashboard from HubSpot and the Q4 budget sheet. Refresh it every morning.

Kluff

Done. Signups come from HubSpot, spend from the Q4 budget sheet. It refreshes at 7:00 AM.

v2 adds the channel split you asked for.

Maya Ruiz

Can we flag any channel over $40 per trial?

Kluff

Added. Paid social is flagged.

Q4 launch dashboard
Q4 launchHubSpot + Q4 budget sheet · refreshed 7:00 AM
Live
Signups1,284+18% vs last week
Trials started31224% of signups
Spend$8,420of $12,000 budget
Daily signups
MonTueWedThuFriSatSunMonTueWed
Send the renewal to Harbor Foods?

It sends from your Gmail, as you, with the draft attached. Nothing else leaves the project.

SENDGmail · Dana Reyes, Harbor Foods purchasing
ATTACHGoogle Drive · Harbor renewal 2026.docx
Finance check done

Ana Chen confirmed net 45 at 9:18 AM. Your rule: every customer email waits for you.

Sent · logged to the project