Benchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Summarize + extract
summarize-extract-030
medium summarizationtranscript
The prompt
user
Summarize the meeting transcript below in at most 80 words. Focus on decisions, owners and deadlines. Reply with only the summary. """ [Weekly marketing sync - Larkspur Bakehouse - transcript, auto-generated] Nadia: Okay, let's start, we've got fifteen minutes. Spring menu launch. Where are we? Owen: So the big thing is the pistachio paste. Our usual supplier, Delmarco, told us yesterday they can't deliver until mid-March. That kills the March 2 date. Nadia: Can we launch without the pistachio croissant? Priscilla: I'd rather not. It's the hero item, it's on every mockup. Owen: There's another option. Harlow Mills can deliver pistachio paste next Wednesday. It's about eight percent more expensive, but Theo already tasted their sample. Theo: It's good. Honestly a little better than Delmarco's. Slightly less sweet. Nadia: Okay. So we switch to Harlow Mills for this batch. Owen, you'll place that order? Owen: Yep, I'll place it today. Nadia: And the launch date? Theo: If the paste arrives Wednesday, the kitchen needs a few days to test. Monday, March 9 is realistic. Nadia: Okay, so we're moving the launch to Monday, March 9. Everyone okay with that? ... Okay. Priscilla, photos? Priscilla: I was going to shoot this Thursday, but I'll need the real croissant, so I'll move the shoot to Tuesday, March 3, at the Elm Street shop. Nadia: Great. Packaging? Owen: The new pastry boxes. I'll get the order to the printer by Friday. They need ten business days. Nadia: Pricing - Theo, did we land on the croissant price? Theo: Four twenty-five. The ingredient cost went up with Harlow, but we think four twenty-five is still fine. Delmarco would've let us do three ninety-five. Nadia: Four twenty-five, done. Last thing - social. I'll write the teaser posts and schedule them for the week before launch. Anything else? No? Thanks everyone. """
What we expected
{
"kind": "summarize",
"word_limit": 80,
"facts": [
{
"id": "supplier",
"desc": "switching pistachio paste supplier to Harlow Mills",
"any": [
"harlow"
]
},
{
"id": "launch",
"desc": "launch moved to Monday March 9",
"any": [
"(\\bmar(ch|\\.)?\\s*9(st|nd|rd|th)?\\b|\\b9(st|nd|rd|th)?\\s+(of\\s+)?mar(ch|\\.)?\\b|\\b0?3/0?9\\b|\\b2026-03-09\\b|\\bninth\\s+(of\\s+)?mar(ch|\\.)?\\b|\\bmar(ch|\\.)?\\s+(the\\s+)?ninth\\b)"
]
},
{
"id": "photos",
"desc": "Priscilla moves photo shoot to Tuesday March 3",
"any": [
[
"priscilla|photo|shoot",
"(\\bmar(ch|\\.)?\\s*3(st|nd|rd|th)?\\b|\\b3(st|nd|rd|th)?\\s+(of\\s+)?mar(ch|\\.)?\\b|\\b0?3/0?3\\b|\\b2026-03-03\\b|\\bthird\\s+(of\\s+)?mar(ch|\\.)?\\b|\\bmar(ch|\\.)?\\s+(the\\s+)?third\\b|\\btue(sday|\\.)?\\b)"
]
]
},
{
"id": "packaging",
"desc": "Owen sends packaging/box order to printer by Friday",
"any": [
[
"owen",
"packag|box|printer"
]
]
},
{
"id": "price",
"desc": "croissant priced at $4.25",
"any": [
"4\\.25|four twenty-five|four twenty five"
]
},
{
"id": "social",
"desc": "Nadia writes/schedules teaser social posts",
"any": [
"nadia.{0,60}(post|social|teaser)|(post|social|teaser).{0,60}nadia"
]
}
],
"allowed_numbers": [
2026
],
"source_words": 283,
"reference_summary": "Delmarco's pistachio paste is delayed, so the team is switching to Harlow Mills; Owen orders it today. The spring launch moves from March 2 to Monday, March 9. Priscilla moves the photo shoot to Tuesday, March 3 at the Elm Street shop. Owen sends the pastry box order to the printer by Friday. The pistachio croissant will cost $4.25. Nadia will write and schedule teaser social posts for the week before launch."
}What each model answered
MiMo-V2.6 Distill 9B
✓ pass · score 100 (scripted checks 100, judges 10.0/10 from 3) · 3.17 s
Decisions: Switch pistachio paste supplier to Harlow Mills (8% costlier); Owen places order today. Launch moves to Monday, March 9; croissant price set at $4.25. Priscilla shoots March 3 at Elm Street shop. Owen orders pastry boxes by Friday (10 business-day lead). Nadia writes and schedules teaser posts for the week before launch.
Judge medians: Faithfulness 10 · Fluency and usefulness 10
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": true,
"social": true
},
"facts_frac": 1,
"word_count": 53,
"word_limit": 80,
"length_score": 1,
"unsupported_numbers": [],
"number_score": 1
}Ornith-1.5 9B
✓ pass · score 100 (scripted checks 100, judges 10.0/10 from 3) · 3.10 s
The team decided to switch pistachio paste suppliers from Delmarco to Harlow Mills due to a mid-March delivery delay. Owen will place the order today. The launch moves from March 2 to Monday, March 9. Priscilla will shoot photos Tuesday, March 3 at Elm Street shop. Owen will order new pastry boxes by Friday. The croissant price is set at $4.25. Nadia will write and schedule teaser social posts for the week before launch.
Judge medians: Faithfulness 10 · Fluency and usefulness 10
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": true,
"social": true
},
"facts_frac": 1,
"word_count": 74,
"word_limit": 80,
"length_score": 1,
"unsupported_numbers": [],
"number_score": 1
}Granite 4.2 8B
✗ fail · score 83 (scripted checks 85, judges 8.0/10 from 3) · 4.18 s
Nadia leads the spring menu launch. Owen will order pistachio paste from Harlow Mills today to replace Delmarco’s delayed supply. Theo confirms the paste is good; kitchen testing will take a few days, making March 9 the realistic launch date. Priscilla will shoot photos on Tuesday, March 3 at the Elm Street shop instead of Thursday. Owen will send pastry‑box order to the printer by Friday so it finishes in ten business days. Theo set the croissant price at $4.25. Nadia will write and schedule teaser social posts for the week before launch. No other actions were mentioned.
Judge medians: Faithfulness 9 · Fluency and usefulness 7
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": true,
"social": true
},
"facts_frac": 1,
"word_count": 98,
"word_limit": 80,
"length_score": 0,
"unsupported_numbers": [],
"number_score": 1
}Spark-X2.5 4B
✗ fail · score 82 (scripted checks 83, judges 8.0/10 from 3) · 1.86 s
Decision: Switch to Harlow Mills for pistachio paste; launch on March 9; shoot photos on March 3; finalize pricing at $4.25. Owner: Owen to place Harlow Mills order today and order new pastry boxes by Friday. Deadline: Photos by March 3; packaging order by Friday.
Judge medians: Faithfulness 10 · Fluency and usefulness 6
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": true,
"social": false
},
"facts_frac": 0.8333,
"word_count": 45,
"word_limit": 80,
"length_score": 1,
"unsupported_numbers": [],
"number_score": 1
}Mellum2.1 12B-A2.5B
✗ fail · score 91 (scripted checks 93, judges 9.0/10 from 3) · 1.88 s
The team decided to switch pistachio paste supplier from Delmarco to Harlow Mills due to delayed delivery, accepting an 8% price increase. Owen will place the order today. The spring menu launch moved from March 2 to Monday, March 9, after kitchen testing. Priscilla will reschedule photo shoot to Tuesday, March 3 at Elm Street shop. Owen will order new pastry boxes by Friday; ten‑business‑day lead time applies. Pricing set at $4.25 per croissant despite higher ingredient cost. Nadia will write teaser posts and schedule them for the week before launch.
Judge medians: Faithfulness 10 · Fluency and usefulness 8
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": true,
"social": true
},
"facts_frac": 1,
"word_count": 91,
"word_limit": 80,
"length_score": 0.5,
"unsupported_numbers": [],
"number_score": 1
}Ling-3.0-tiny (7.9B-A1.3B)
✗ fail · score 82 (scripted checks 83, judges 8.0/10 from 3) · 1.10 s
The marketing team decided to switch to Harlow Mills for pistachio paste, placing the order today. The launch date is moved to Monday, March 9. Owen will order new pastry boxes by Friday, with a 10-business-day delivery. Priscilla will photograph the croissant at the Elm Street shop on Tuesday, March 3. The croissant price is set at $4.25. Social media teaser posts will be written and scheduled for the week before launch.
Judge medians: Faithfulness 9 · Fluency and usefulness 7
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": true,
"social": false
},
"facts_frac": 0.8333,
"word_count": 72,
"word_limit": 80,
"length_score": 1,
"unsupported_numbers": [],
"number_score": 1
}MiniCPM5 2B
✗ fail · score 64 (scripted checks 77, judges 5.0/10 from 3) · 1.25 s
Decisions: Switch to Harlow Mills for pistachio paste, move launch to March 9, adjust photo shoot to March 3, order pastry boxes by Friday, confirm croissant price at $42.25. Owners: Owen (order, printer), Theo (price), Nadia (social, launch). Deadlines: Order today, printer by Friday, shoot by March 3.
Judge medians: Faithfulness 5 · Fluency and usefulness 5
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": false,
"social": true
},
"facts_frac": 0.8333,
"word_count": 48,
"word_limit": 80,
"length_score": 1,
"unsupported_numbers": [
"42.25"
],
"number_score": 0.5
}K2-Horizon 0.9B
✗ fail · score 53 (scripted checks 67, judges 4.0/10 from 3) · 0.50 s
Theo switches to Harlow Mills for pistachio paste, launching March 9 instead of March 2. Nadia approves the change; Owen places the order; Theo confirms paste delivery Wednesday; pricing stays at $25; photos move to Tuesday; packaging order by Friday; social posts scheduled for week before launch.
Judge medians: Faithfulness 3 · Fluency and usefulness 5
Scoring detail
{
"facts": {
"supplier": true,
"launch": true,
"photos": true,
"packaging": true,
"price": false,
"social": false
},
"facts_frac": 0.6667,
"word_count": 47,
"word_limit": 80,
"length_score": 1,
"unsupported_numbers": [],
"number_score": 1
}← summarize-extract-029All Summarize + extract itemssummarize-extract-031 →