Start free
Read & reportRead-only source

See your Replicate numbers in AWS S3.

Review Replicate prediction statuses, predict times and dedicated deployments in a read-only data source.

  • Traffic, sources and conversions in a sheet or dashboard, refreshed for you
  • A weekly traffic report nobody has to rebuild
  • Ask ChatGPT or Claude what changed, using live Replicate numbers
ReplicateAWS S3AWS S3

Start free Connect Replicate

Updated November 2026

Replicate reportSample data
ModelStatusSource
black-forest-labs/flux-1.1-prosucceeded1,240
640Pending640
210Refunded210
Refreshes on your schedule in AWS S3

Who uses Datapipe with Replicate

Marketers

See which channels and pages bring visitors who convert.

Founders

Get a weekly traffic summary without opening another dashboard.

Agencies

Send each client the same clean traffic report every month.

What you can do with Replicate

Read and report

Read predictions, deployments and account details using your Replicate API token.

  • Model predictions
  • Deployments
  • Account info

Explore metrics and dimensions

Read-only source

Datapipe reads Replicate data for reports. It never changes anything in Replicate.

Full read details

Models and predictions are never modified.

Examples for your team

Read & report

Monitor AI model prediction latency and deployments

Combine prediction status, runtime seconds and deployment releases in one report.

  1. Scheduled refresh
  2. Latest Replicate data
  3. Updated report

Before you connect

What you need

An API key or token from your Replicate account.

Read-only

Read access only. Datapipe never changes anything in Replicate.

Limits and exclusions

Read-only. No prediction is run, inputs and outputs are not included, and predict time is in seconds.

Full details and limits

AI model predictions, predict durations, and deployments from Replicate.

Read-only reporting access using your Replicate API token. Models and predictions are never modified.

Read-only. No prediction is run, inputs and outputs are not included, and predict time is in seconds.

Reports you can build

3 reporting datasets, each with its own fields. Pick one in the app, then choose your columns.

What each dataset reports
  • Model predictions

    Replicate predictions created in the report dates with model, status, source, predict time in seconds, and creation date.

  • Deployments

    Replicate dedicated model deployments with owner, name, and active release model.

  • Account info

    Replicate account type, username, display name, and GitHub URL.

Metrics and dimensions

14 report fields across 3 datasets.

Hover or tap a field to see what it means. Custom fields from your own account appear after you connect.

Send Replicate data anywhere

Ask AI about your data

Claude or ChatGPT can read the same numbers. You ask in a normal sentence, and the answer stays with your report.

Summarize Replicate prediction volumes, predict times, and deployments.

flux-1.1-pro ran 3,420 predictions with an average predict time of 4.2 seconds.

Better than exporting by hand

Start from a goal

How it works

  1. Create a project and choose Replicate in Data sources.

  2. Pick where it goes. This one sends the numbers to AWS S3.

  3. Pick where it goes. Choose predictions, deployments or the account and a destination.

Sample report

Replicate reportSample data
ModelStatusSourcePredict time (s)Created atOwnerDeployment nameRelease model
black-forest-labs/flux-1.1-prosucceeded1,2404.22026-11-131,2401,2401,240
640Pending640640640640640640
210Refunded210210210210210210

Sample numbers, so you can see the shape. Your report uses your accounts.

Works well with Replicate

Add both to the same project. Each keeps its own report, so you can check them side by side or ask ChatGPT or Claude a question that needs both.

Questions

Free to start. Paid plans add more data sources and refresh faster.

Free and Starter refresh once a day. Pro refreshes every hour. Agency refreshes every 15 minutes on the first 50 data sources.

No. Name a project, add a data source, then choose a report and destination.

OAuth means you sign in with the service and grant access, without pasting an API key. Datapipe works with multiple partners to provide that sign-in. Some partners already provide it. Some integrations are API-only at present, so you paste a key or token. OAuth is coming soon for some. Some partners do not provide OAuth at all. The setup steps on each integration page say how that source connects. Signing in is only how the connection is made. It is not permission to change records. Where a source supports changes, those writes stay off until the workspace owner turns specific actions on and approves each one. We do not promise OAuth on every source, and we do not share dates.

This integration is read-only in Datapipe. You can report on Replicate, send the reports to your destinations and ask an AI assistant about them. Datapipe never changes anything in Replicate. Read-only reporting access using your Replicate API token. Models and predictions are never modified.

This page lists 14 report fields across 3 datasets. You pick the columns when you build the report.

AI model predictions, predict durations, and deployments from Replicate. Account permissions and the records available from the provider determine which rows can be read.

One for each Replicate account you add to a project. Adding the same account to another project counts again. Reports and destinations do not use extra data sources.

Yes. The same numbers can go to a sheet, a chart, a warehouse, or an assistant.

Try Replicate with sample data first

Build the report first, then add Replicate when you are ready.

Start free Add data source