PUBLIC METHOD · VERSION 1.0

How LocalLens tests local AI.

A result is trustworthy only when another person can understand the setup, repeat the test, and see what failed.

Submit a test result
Same setup

Save the model, engine, settings, and hardware.

Same task

Do not change the prompt between models.

Visible failure

Wrong answers and crashes stay in the record.

Clear evidence

Labels show what LocalLens has and has not checked.

1
STEP 1

Record the computer

  • Operating system and version
  • Exact CPU
  • Exact GPU and dedicated VRAM
  • Total RAM
  • Free storage
2
STEP 2

Record the software

  • Engine and version
  • Exact model tag or file
  • Quantization
  • Context setting
  • Driver version when relevant
3
STEP 3

Prepare the test

  • Use a real task with private details removed
  • Save the exact prompt
  • Write the expected result before testing
  • Include a question the model should not answer when useful
4
STEP 4

Run the model

  • Run one warm-up prompt
  • Use the saved prompt without changing it
  • Record the time before the first word
  • Record speed and highest memory when possible
  • Save failures and error messages
5
STEP 5

Judge the result

  • Was the main answer correct?
  • Did it follow the requested format?
  • Did it invent facts?
  • Was the waiting time useful?
  • Could another person repeat the setup?
EVIDENCE LABELS

What each label means

Submitted

A user sent the result. LocalLens has not confirmed the setup.

Community confirmed

More than one independent report supports the same useful result.

LocalLens verified

The LocalLens team reviewed the evidence and repeated or directly checked the important parts.

Have a result another person can repeat?

Contribute your test