Claude / Eval / Harness

Model
comparison grid

Send one prompt to multiple Claude models. Compare outputs, latency, and cost. Your key stays in this tab.

01

API key

Get a key at console.anthropic.com. Held in this browser tab only. Never written to localStorage, sessionStorage, cookies, or any URL. Closing or refreshing this tab discards it.

OK Key validated. Stored in memory for this session only.
02

Prompt

The same prompt will be sent to every selected model. Cmd/Ctrl+Enter runs from inside this field.

Advanced options

Higher = longer outputs allowed. Higher = costs more if used.

03

Models

Select one or more models. Pricing shown is per 1M tokens (input / output). Verify current pricing at anthropic.com/pricing before relying on cost numbers.

04

Results