Start free

Helicone connector · LLM observability reports

Read & reportRead-only source

See what every LLM request costs in Helicone.

Bring Helicone requests, daily cost and per-user spend into your reports and AI assistants, with model, provider, latency, tokens and cost for each request.

  • Track daily LLM spend across providers
  • Find slow or failing requests by model
  • Ask ChatGPT or Claude which users cost the most
HeliconeSheetsSheetsData StudioData StudioChatGPTChatGPTClaudeClaude

Start free Explore the data

Updated November 2026

Helicone reportSample data
ModelProviderLatency (ms)
gpt-4.1-miniOPENAI842
claude-sonnet-4-5ANTHROPIC2,316
gpt-4.1OPENAI1,904
Refreshes on your schedule in Google Sheets

Helicone reporting for one organization

One Helicone organization is the data source. All three reports follow your report dates.

Requests

Read logged requests with model, provider, HTTP status, latency, prompt and completion tokens and cost.

Daily cost

See total LLM spend for each day.

Users

Compare requests, tokens and spend per user, with when each was last active.

Errors

Count failed requests by model and provider.

Report datasets and details

Reports you can build

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

  • LLM requests
  • Daily LLM cost
  • User metrics
What each dataset reports
  • LLM requests

    LLM requests logged in Helicone in the report period with model, provider, HTTP status, latency in milliseconds, prompt and completion tokens, and cost in USD.

  • Daily LLM cost

    Helicone daily aggregated LLM spend over time.

  • User metrics

    Helicone per-user request volume, total tokens, and spend.

Metrics and dimensions

13 report fields across 3 datasets.

Search the available fields. Hover or tap a field to see what it means.

Send Helicone data anywhere

Connect Helicone to ChatGPT or Claude for LLM cost questions

Use Datapipe’s MCP access as your Helicone AI integration. Ask about spend, latency and errors.

Example

Which models had the slowest Helicone requests last week?

Average last week’s request latency by model, then sort from slowest to fastest and check their error counts.

Start with Helicone

What you need

A Helicone API key from your Helicone settings. The US or EU region is detected from the key.

What to know

Read-only. Datapipe never sends a request through Helicone. Prompts and responses aren’t included. One report reads up to 20,000 requests.

How it works

  1. Create your free account, then choose Helicone in Data sources.

  2. Paste your Helicone API key.

  3. Choose a report and destination: a sheet, Data Studio, ChatGPT or Claude.

Works well with Helicone

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

Datapipe reads logged requests, daily cost and per-user metrics for your report dates.

No. Reports cover request details, tokens and cost, not the text.

Yes. The US or EU region is detected from your key.

One report reads up to 20,000 requests. Use shorter report dates for busy projects.

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.

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

Keep LLM costs in view.

Start free, add Helicone and bring its requests into the tools you already use.

Start free See setup requirements