LLM Classification Tool
Upload a dataset, configure your model, and run classifications with live monitoring.
How to Use
Toggle
Pick a test case or upload your own API key, prompt, schema, and CSV.
Map the ID and text columns and optionally review/edit rows.
Choose a model, estimate cost, then start the run.
Review results, charts, and download JSONL.
Test Cases
Toggle
Use Test Case
Select a test case...
L1 Level (Subsample)
P1 Promise (Sample)
Load Test Case
Loading a test case replaces the current prompt, schema, and dataset in memory.
Inputs
Toggle
API Key
Kept in-memory only for this session.
Drop API key file or
Browse
System Prompt
Drop prompt file or
Browse
JSON Schema / response_format
Paste a full response_format object or a raw schema with "properties".
Drop schema file or
Browse
Dataset
Toggle
CSV File
CSV should include an ID column and an input text column.
ID Column
Text Column
Review Inputs
Toggle
Prev
Row
0
of
0
Next
Remove Row
ID
Text
Edits update the in-memory dataset used for classification.
Model Settings
Toggle
Model
gpt-4o
gpt-4o-mini
gpt-4-turbo
gpt-3.5-turbo
Currently supports gpt-4, gpt-3.5, gpt-4o, and gpt-4o-mini variants.
Temperature
Top Logprobs
None
5
10
20
Seed (Optional)
Target RPM
Estimate Cost
Input Tokens
0
Output Tokens
0
Total Cost
$0.00
Run
Toggle
Start Classification
Stop
Download Results
Progress
0 / 0
Errors
0
Concurrency
0
Effective RPM
0
Results
Toggle
Model
N/A
Fingerprints
0
Total Time
0s
Total Cost
$0.0000
Status
ID
Class
Top-1
Margin
Time
Cost
Info
Activity
Toggle
Log