Unified Memory

Benchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Read images

vision-032

hard handwritingcounting

The prompt

1 image attached (not shown).

user

On this shopping list, how many items are NOT crossed out? Reply with only the number (no units or currency symbols).
What we expected
{
  "answer": 6,
  "accept": [],
  "type": "number"
}

What each model answered

LFM2.5-VL 3B

✗ fail · score 0 · 2.91 s

4
Scoring detail
{
  "type": "number",
  "parsed": 4
}

MiMo-V2.6 Distill 9B

✗ fail · score 0 · 2.45 s

7
Scoring detail
{
  "type": "number",
  "parsed": 7
}

MiniCPM-V 4.7 1B

✗ fail · score 0 · 1.20 s

5
Scoring detail
{
  "type": "number",
  "parsed": 5
}

← vision-031All Read images itemsvision-033 →