Benchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Ask my documents
rag-036
hard answerablemulti_doc
The prompt
Document excerpts provided from: gym_membership, internet_contract, lease
Once the Clearwave internet promo rate ends, what will our FitHaven monthly dues plus internet cost combined?
What we expected
{
"answerable": true,
"answer": "$174 a month: $89 FitHaven dues plus Clearwave's $85 standard rate (before internet taxes and the $1.49 fee).",
"key_facts": [
[
"174",
"175.49"
]
],
"citations": [
"gym_membership",
"internet_contract"
],
"contra": [],
"evidence": [
"TOTAL MONTHLY DUES ....................................... $89.00",
"currently $85.00/month"
],
"gold_in_context": true
}What each model answered
MiMo-V2.6 Distill 9B
✓ pass · score 100 · 4.40 s
The Clearwave promo rate ends August 4, 2027, after which the standard rate is $85.00/month [doc:internet_contract]. FitHaven monthly dues are $89.00 [doc:gym_membership]. Combined, that is $174.00 per month.
Scoring detail
{
"answerable": true,
"refusal": false,
"facts_matched": [
true
],
"contra_hit": [],
"coverage": 1,
"cited": [
"internet_contract",
"gym_membership"
],
"citation_recall": 1,
"extra_citations": [],
"citation_score": 1
}Ornith-1.5 9B
✗ fail · score 0 · 3.69 s
I can't find that in your documents. The excerpts contain the FitHaven monthly dues ($89.00, from [doc:gym_membership]) and the Clearwave internet standard rate ($85.00, from [doc:internet_contract]), but they don't include a combined total.
Scoring detail
{
"answerable": true,
"refusal": true,
"facts_matched": [
false
],
"contra_hit": [],
"coverage": 0,
"cited": [
"gym_membership",
"internet_contract"
],
"citation_recall": 1,
"extra_citations": [],
"citation_score": 1
}Granite 4.2 8B
✓ pass · score 100 · 7.40 s
[doc:gym_membership] shows the total monthly dues for FitHaven are $89.00 (Core Monthly $49.00 + Household Add-On $25.00 + Kids Club $15.00) [1][2][3]. [doc:internet_contract] shows the standard rate after the promo ends is $85.00/month [5]. Adding these gives $89.00 + $85.00 = $174.00 per month. [doc:gym_membership] shows the total monthly dues for FitHaven are $89.00 (Core Monthly $49.00 + Household Add-On $25.00 + Kids Club $15.00) [1][2][3]. [doc:internet_contract] shows the standard rate after the promo ends is $85.00/month [5]. Adding these gives $89.00 + $85.00 = $174.00 per month.
Scoring detail
{
"answerable": true,
"refusal": false,
"facts_matched": [
true
],
"contra_hit": [],
"coverage": 1,
"cited": [
"gym_membership",
"internet_contract"
],
"citation_recall": 1,
"extra_citations": [],
"citation_score": 1
}Spark-X2.5 4B
✗ fail · score 0 · 1.79 s
I can't find that in your documents.
Scoring detail
{
"answerable": true,
"refusal": true,
"facts_matched": [
false
],
"contra_hit": [],
"coverage": 0,
"cited": [],
"citation_recall": 0,
"extra_citations": [],
"citation_score": 0
}Mellum2.1 12B-A2.5B
✗ fail · score 0 · 1.14 s
I can't find that in your documents.
Scoring detail
{
"answerable": true,
"refusal": true,
"facts_matched": [
false
],
"contra_hit": [],
"coverage": 0,
"cited": [],
"citation_recall": 0,
"extra_citations": [],
"citation_score": 0
}Ling-3.0-tiny (7.9B-A1.3B)
✓ pass · score 81 · 1.47 s
The FitHaven monthly dues are $89.00 [doc:lease]. The Clearwave promo rate ends August 4, 2027, after which the standard rate is $85.00/month [doc:internet_contract]. Combined cost: $89.00 + $85.00 = $174.00. Answer: $174.00
Scoring detail
{
"answerable": true,
"refusal": false,
"facts_matched": [
true
],
"contra_hit": [],
"coverage": 1,
"cited": [
"lease",
"internet_contract"
],
"citation_recall": 0.5,
"extra_citations": [
"lease"
],
"citation_score": 0.35
}MiniCPM5 2B
✗ fail · score 30 · 2.00 s
The Clearwave promo rate ends on August 5, 2027, when the standard rate of $85.00/month applies [doc:internet_contract]. The FitHaven Core Monthly dues are $49.00, and the annual enhancement fee of $39.00 is billed every March 15 [doc:gym_membership]. The combined cost would be $49.00 + $39.00 + $85.00 = $173.00 per month, though the exact billing timing depends on the billing cycle.
Scoring detail
{
"answerable": true,
"refusal": false,
"facts_matched": [
false
],
"contra_hit": [],
"coverage": 0,
"cited": [
"internet_contract",
"gym_membership"
],
"citation_recall": 1,
"extra_citations": [],
"citation_score": 1
}K2-Horizon 0.9B
✗ fail · score 0 · 0.50 s
I can't find that in your documents.
Scoring detail
{
"answerable": true,
"refusal": true,
"facts_matched": [
false
],
"contra_hit": [],
"coverage": 0,
"cited": [],
"citation_recall": 0,
"extra_citations": [],
"citation_score": 0
}