Benchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Rewrite and proofread

rewrite-017

easy tonecasual-to-professional

The prompt

user

Rewrite the message below in a clear, professional tone suitable for a work email. Keep every fact (names, dates, times, amounts, reference numbers) as given, and do not add new facts or numbers. Reply with only the rewritten message, with no preamble or explanation.

"""
hey!! so the logo files r ready lol. sending them over tmrw morning, gonna be 3 versions (color, black, white). invoice is $650, due in 2 weeks. lmk if u want any tweaks before then
"""

What we expected

{
  "type": "tone",
  "style": "casual-to-professional",
  "source": "hey!! so the logo files r ready lol. sending them over tmrw morning, gonna be 3 versions (color, black, white). invoice is $650, due in 2 weeks. lmk if u want any tweaks before then",
  "facts": [
    {
      "name": "logo",
      "any": [
        "logo"
      ]
    },
    {
      "name": "tomorrow morning",
      "any": [
        "tomorrow morning",
        "tmrw morning"
      ]
    },
    {
      "name": "3 versions",
      "any": [
        "3",
        "three"
      ]
    },
    {
      "name": "color",
      "any": [
        "color",
        "colour"
      ]
    },
    {
      "name": "black",
      "any": [
        "black"
      ]
    },
    {
      "name": "white",
      "any": [
        "white"
      ]
    },
    {
      "name": "$650",
      "any": [
        "$650",
        "650 dollars"
      ]
    },
    {
      "name": "2 weeks",
      "any": [
        "2 weeks",
        "two weeks"
      ]
    }
  ],
  "banned": [
    "lol",
    "lmk",
    "u",
    "r",
    "tmrw",
    "gonna",
    "!!"
  ],
  "banned_re": [],
  "no_caps": false,
  "caps_ok": [],
  "extra_numbers_ok": [],
  "length_ratio": [
    0.5,
    3
  ],
  "reference": "Hello,\n\nThe logo files are ready. I will send them over tomorrow morning in 3 versions: color, black, and white. The invoice is $650, due in 2 weeks. Please let me know if you would like any adjustments before then.\n\nBest regards"
}

What each model answered

MiMo-V2.6 Distill 9B

✓ pass · score 98 (scripted checks 100, judges 9.7/10 from 3) · 2.26 s

Hi,

The logo files are ready. I'll send them over tomorrow morning, and there will be three versions: color, black, and white. The invoice is $650 and is due in two weeks. Please let me know if you'd like any tweaks before then.

Best,
[Your Name]

Judge medians: Tone match 10 · Faithfulness 10 · Naturalness 9

Scoring detail

{
  "cleaned_words": 46,
  "similarity_to_source": 0.42,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.314,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

Ornith-1.5 9B

✓ pass · score 97 (scripted checks 100, judges 9.3/10 from 3) · 1.80 s

Hi,

The logo files are ready. I'll be sending them over tomorrow morning, and there will be three versions: color, black, and white. The invoice is $650 and is due in two weeks. Please let me know if you'd like any tweaks before then.

Judge medians: Tone match 9 · Faithfulness 10 · Naturalness 9

Scoring detail

{
  "cleaned_words": 44,
  "similarity_to_source": 0.456,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.257,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

Granite 4.2 8B

✓ pass · score 97 (scripted checks 100, judges 9.3/10 from 3) · 2.31 s

Subject: Logo Files and Invoice

Dear [Recipient],

The logo files are ready. I will send them over tomorrow morning, and they will include three versions (color, black, and white). The invoice amount is $650 and is due in two weeks. Please let me know if you would like any tweaks before then.

Best regards,  
[Your Name]

Judge medians: Tone match 10 · Faithfulness 10 · Naturalness 8

Scoring detail

{
  "cleaned_words": 56,
  "similarity_to_source": 0.418,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.6,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

Spark-X2.5 4B

✓ pass · score 99 (scripted checks 100, judges 9.8/10 from 2) · 1.21 s

Hi,

The logo files are ready and will be sent to you tomorrow morning. There are three versions included: color, black, and white. The invoice is for $650 and is due in two weeks. Please let me know if you would like any adjustments before then.

Judge medians: Tone match 10 · Faithfulness 10 · Naturalness 9.5

Scoring detail

{
  "cleaned_words": 46,
  "similarity_to_source": 0.37,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.314,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

Mellum2.1 12B-A2.5B

✓ pass · score 98 (scripted checks 100, judges 9.7/10 from 3) · 0.77 s

The logo files are ready. I will send them over tomorrow morning. There will be three versions: color, black, and white. The invoice amounts to $650 and is due within two weeks. Please let me know if you would like any tweaks before then.

Judge medians: Tone match 10 · Faithfulness 10 · Naturalness 9

Scoring detail

{
  "cleaned_words": 44,
  "similarity_to_source": 0.38,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.257,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

Ling-3.0-tiny (7.9B-A1.3B)

✓ pass · score 98 (scripted checks 100, judges 9.7/10 from 3) · 0.61 s

Hi, the logo files are ready. I will be sending them over tomorrow morning, and they will include three versions (color, black, and white). The invoice is $650, due in 2 weeks. Please let me know if you would like any adjustments before that date.

Judge medians: Tone match 10 · Faithfulness 10 · Naturalness 9

Scoring detail

{
  "cleaned_words": 45,
  "similarity_to_source": 0.525,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.286,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

MiniCPM5 2B

✓ pass · score 93 (scripted checks 100, judges 8.7/10 from 3) · 0.90 s

The logo files are ready. They will be sent tomorrow morning. There are three versions available: color, black, and white. The invoice is $650, due in two weeks. Please let me know if you need any adjustments before then.

Judge medians: Tone match 9 · Faithfulness 10 · Naturalness 7

Scoring detail

{
  "cleaned_words": 39,
  "similarity_to_source": 0.405,
  "facts_missing": [],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 1.114,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 1,
  "checks": {
    "facts": 1,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

K2-Horizon 0.9B

✗ fail · score 43 (scripted checks 33, judges 5.3/10 from 3) · 0.22 s

Hi team,

The logo files are ready for tomorrow morning. We will send them over the next few hours. Please let me know if you need any adjustments before then.

Judge medians: Tone match 9 · Faithfulness 2 · Naturalness 5

Scoring detail

{
  "cleaned_words": 30,
  "similarity_to_source": 0.277,
  "facts_missing": [
    "3 versions",
    "color",
    "black",
    "white",
    "$650",
    "2 weeks"
  ],
  "banned_hits": [],
  "new_numbers": [],
  "length_ratio": 0.857,
  "length_bounds": [
    0.5,
    3
  ],
  "gate": 0.5,
  "checks": {
    "facts": 0.25,
    "banned": 1,
    "numbers": 1,
    "changed": 1,
    "length": 1
  }
}

← rewrite-016All Rewrite and proofread itemsrewrite-018 →