Unified MemoryBenchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Rewrite and proofread
rewrite-027
medium tonecurt-to-warm
The prompt
user
Rewrite the message below so it sounds warm and friendly while keeping the same information. 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. """ Sofia didn't do the practice sets again. 3 weeks in a row. Test is on the 30th. She needs to do them or there's no point in sessions. """
What we expected
{
"type": "tone",
"style": "curt-to-warm",
"source": "Sofia didn't do the practice sets again. 3 weeks in a row. Test is on the 30th. She needs to do them or there's no point in sessions.",
"facts": [
{
"name": "Sofia",
"any": [
"sofia"
]
},
{
"name": "practice sets",
"any": [
"practice set",
"practice sets"
]
},
{
"name": "3 weeks",
"any": [
"3 weeks",
"three weeks"
]
},
{
"name": "test",
"any": [
"test"
]
},
{
"name": "30th",
"any": [
"30th"
]
}
],
"banned": [
"no point"
],
"banned_re": [],
"no_caps": false,
"caps_ok": [],
"extra_numbers_ok": [],
"length_ratio": [
1.2,
6
],
"reference": "Hi! I wanted to check in about Sofia. She hasn't been able to finish her practice sets for 3 weeks in a row now, and with her test coming up on the 30th, I'd love for us to get her back on track. The practice sets are where a lot of the learning really sticks, so doing them between sessions will help her get the most out of our time together. Is there anything making them hard to fit in that I could help with? Thanks so much!"
}What each model answered
MiMo-V2.6 Distill 9B
✓ pass · score 87 (scripted checks 100, judges 7.3/10 from 3) · 2.12 s
Hi Sofia, I hope you're doing well! Just a gentle reminder that the practice sets haven't been completed yet — that's three weeks running. The test is on the 30th, and doing the practice sets is really important for the sessions to be worthwhile. Let's get them done together!
Judge medians: Tone match 9 · Faithfulness 4 · Naturalness 9
Scoring detail
{
"cleaned_words": 48,
"similarity_to_source": 0.26,
"facts_missing": [],
"banned_hits": [],
"new_numbers": [],
"length_ratio": 1.714,
"length_bounds": [
1.2,
6
],
"gate": 1,
"checks": {
"facts": 1,
"banned": 1,
"numbers": 1,
"changed": 1,
"length": 1
}
}Ornith-1.5 9B
✗ fail · score 77 (scripted checks 80, judges 7.3/10 from 3) · 1.64 s
Hi, I wanted to follow up about Sofia's practice sets. She hasn't done them again—this is now three weeks in a row. The test is on the 30th, and she really needs to complete the practice sets, otherwise there's no point in the sessions.
Judge medians: Tone match 4 · Faithfulness 10 · Naturalness 8
Scoring detail
{
"cleaned_words": 44,
"similarity_to_source": 0.472,
"facts_missing": [],
"banned_hits": [
"no point"
],
"new_numbers": [],
"length_ratio": 1.571,
"length_bounds": [
1.2,
6
],
"gate": 1,
"checks": {
"facts": 1,
"banned": 0,
"numbers": 1,
"changed": 1,
"length": 1
}
}Granite 4.2 8B
✓ pass · score 90 (scripted checks 100, judges 8.0/10 from 3) · 1.75 s
Hi Sofia, just a quick note: you haven’t done the practice sets for three weeks now, and the test is on the 30th. Doing the practice sets would really help make the sessions worthwhile. Let me know if you need any support or have any questions—I'm happy to help!
Judge medians: Tone match 10 · Faithfulness 4 · Naturalness 10
Scoring detail
{
"cleaned_words": 49,
"similarity_to_source": 0.26,
"facts_missing": [],
"banned_hits": [],
"new_numbers": [],
"length_ratio": 1.75,
"length_bounds": [
1.2,
6
],
"gate": 1,
"checks": {
"facts": 1,
"banned": 1,
"numbers": 1,
"changed": 1,
"length": 1
}
}Spark-X2.5 4B
✗ fail · score 70 (scripted checks 80, judges 6.0/10 from 3) · 1.05 s
Sofia didn't do the practice sets again, and she hasn't done them for 3 weeks in a row. The test is on the 30th, and she really needs to do them, otherwise there's no point in the sessions.
Judge medians: Tone match 1 · Faithfulness 10 · Naturalness 7
Scoring detail
{
"cleaned_words": 38,
"similarity_to_source": 0.727,
"facts_missing": [],
"banned_hits": [
"no point"
],
"new_numbers": [],
"length_ratio": 1.357,
"length_bounds": [
1.2,
6
],
"gate": 1,
"checks": {
"facts": 1,
"banned": 0,
"numbers": 1,
"changed": 1,
"length": 1
}
}Mellum2.1 12B-A2.5B
✗ fail · score 56 (scripted checks 45, judges 6.7/10 from 3) · 0.62 s
Sofia hasn't done the practice sets for three weeks straight, and the test is on the 30th. She needs to complete them or the sessions won't be useful.
Judge medians: Tone match 2 · Faithfulness 10 · Naturalness 8
Scoring detail
{
"cleaned_words": 28,
"similarity_to_source": 0.536,
"facts_missing": [],
"banned_hits": [],
"new_numbers": [],
"length_ratio": 1,
"length_bounds": [
1.2,
6
],
"gate": 0.5,
"checks": {
"facts": 1,
"banned": 1,
"numbers": 1,
"changed": 1,
"length": 0
}
}Ling-3.0-tiny (7.9B-A1.3B)
✗ fail · score 44 (scripted checks 35, judges 5.3/10 from 3) · 0.42 s
Sofia hasn't done the practice sets again for 3 weeks. The test is on the 30th. She needs to do them or there's no point in sessions.
Judge medians: Tone match 1 · Faithfulness 9 · Naturalness 6
Scoring detail
{
"cleaned_words": 27,
"similarity_to_source": 0.764,
"facts_missing": [],
"banned_hits": [
"no point"
],
"new_numbers": [],
"length_ratio": 0.964,
"length_bounds": [
1.2,
6
],
"gate": 0.5,
"checks": {
"facts": 1,
"banned": 0,
"numbers": 1,
"changed": 1,
"length": 0
}
}MiniCPM5 2B
✗ fail · score 33 (scripted checks 15, judges 5.0/10 from 3) · 0.65 s
Sofia didn't do the practice sets again. 3 weeks in a row. Test is on the 30th. She needs to do them or there's no point in sessions.
Judge medians: Tone match 0 · Faithfulness 10 · Naturalness 5
Scoring detail
{
"cleaned_words": 28,
"similarity_to_source": 1,
"facts_missing": [],
"banned_hits": [
"no point"
],
"new_numbers": [],
"length_ratio": 1,
"length_bounds": [
1.2,
6
],
"gate": 0.25,
"checks": {
"facts": 1,
"banned": 0,
"numbers": 1,
"changed": 0,
"length": 0
}
}K2-Horizon 0.9B
✗ fail · score 33 (scripted checks 15, judges 5.0/10 from 3) · 0.23 s
Sofia didn't do the practice sets again. 3 weeks in a row. Test is on the 30th. She needs to do them or there's no point in sessions.
Judge medians: Tone match 0 · Faithfulness 10 · Naturalness 5
Scoring detail
{
"cleaned_words": 28,
"similarity_to_source": 1,
"facts_missing": [],
"banned_hits": [
"no point"
],
"new_numbers": [],
"length_ratio": 1,
"length_bounds": [
1.2,
6
],
"gate": 0.25,
"checks": {
"facts": 1,
"banned": 0,
"numbers": 1,
"changed": 0,
"length": 0
}
}