AigeoRadar

AigeoRadar · AI Context Benchmark

AI Context Benchmark

Question → sufficient layer(s) → retrieved slice → measured cost → evidence

Bilkent Kanal Açma

https://www.ankaratesisatservisi.com/bilkent-kanal-acma/

20 questions Answer LLM: gpt-4o-mini ai_context_benchmark_v1

Of 20 questions: lowest measured retrieval cost among sufficient layers — HTML 9 · Schema 3 · AIPM 5 · multi-layer 0 · unanswered 3. Descriptive counts only; no overall ranking.

Of 20 questions: lowest measured retrieval cost among sufficient layers — HTML 9 · Schema 3 · AIPM 5 · multi-layer 0 · unanswered 3. Descriptive counts only; no overall ranking. Each question shows: sufficient layer(s) → retrieved slice → measured input tokens → evidence. Readers interpret; the report does not rank formats. Semantic Redundancy 50% — the manifest repeats the same ideas across fields. This inflates structural size without helping per-question retrieval.

Gold mode · Independent HTML gold

Gold answers are derived from the HTML page (title, h1, lang, prose) — not from AIPM fields. Understanding/Retrieval therefore measure layer capability, not AIPM self-consistency. Human-authored gold.json remains the gold standard for publication-grade claims.

Execution provenance

Methodology
AI Context Benchmark Methodology v1.0
Planner Spec
v1.0
Execution Protocol
v1.0
Question pack
universal_v1
Score version
aipm_benchmark_score_v10
Engine
aipm_benchmark_v4
Answer LLM
openai / gpt-4o-mini (single provider this lab)
Model snapshot
2026-07
Run id
9180794c-4d57-4ebf-8efc-d16ad84b5cea

Question outcomes

Per question: which layers were sufficient, and what was the lowest measured retrieval cost among them. No overall winner.

9

Lowest cost: HTML

3

Lowest cost: Schema

5

Lowest cost: AIPM

0

Multi-layer

3

Unanswered

Manifest design

Semantic Redundancy · 50%

Semantic Redundancy 50% — the manifest repeats the same ideas across fields. This inflates structural size without helping per-question retrieval.

  • →Merge or differentiate `purpose` and `abstract` (100% overlap).
Field A Field B Overlap
purpose abstract 100%
purpose keyFacts 26%
abstract keyFacts 26%

Structural size (secondary)

Structural size is secondary. A larger AIPM is not a failure if per-question retrieval stays tiny — check Semantic Redundancy instead.

HTML · 7,126 chars
Schema · 10,757 chars
AIPM · 12,198 chars

Routing helper (secondary)

Illustrative card-first vs HTML-always simulation — prefer per-question measured cost above. Not a ranking.

Card-first retrieval would use ~6% fewer tokens than HTML-always on this pack (11 answered from machine card, 9 escalated to HTML). Descriptive only.

Orientation pack

10 questions · all layers scored

  • HTML 9/10
  • Schema 7/10
  • AIPM 7/10

Depth pack

10 questions · all layers scored

  • HTML 7/10
  • Schema 4/10
  • AIPM 4/10

Full-context pack metrics (secondary)

These measure the whole file fed to the model this run — not the minimum slice needed per question.

HTML

16/20 matched

7,308 full-pack tokens

Schema

11/20 matched

8,144 full-pack tokens

AIPM

11/20 matched

8,287 full-pack tokens

Full-pack resource table (secondary)

Whole-file context fed this run. Prefer Minimal Retrieval Cost on each question.

Metric HTML Schema AIPM
Coverage (matched) 16/20 11/20 11/20
Context size 7,126 chars 10,757 chars 12,198 chars
Total tokens 7,308 8,144 8,287
Tokens / correct answer 457 740 753
Est. cost / correct answer $0.000077 $0.000120 $0.000123
Matched per 1k tokens 2.189 1.351 1.327
Median latency 1,899 ms 2,128 ms 2,066 ms
Est. cost (USD) $0.00124 $0.00132 $0.00135

Six independent scores

Answer Efficiency is the primary cost lens. Accuracy axes remain for research — no combined total or winner.

Answer Efficiency

Matched answers per 1k tokens (and cost per match). The primary efficiency axis — not raw accuracy.

  • HTML 100
  • Schema 61.7
  • AIPM 60.6

Understanding

Can this layer convey what the page is about — using independent HTML gold?

  • HTML 85.7
  • Schema 85.7
  • AIPM 71.4

Retrieval

Can this layer surface shared facts (location, contact, hours, pricing, FAQ, CTA)?

  • HTML 77.8
  • Schema 44.4
  • AIPM 44.4

Evidence

Answer quality vs independent gold (score strength). Partial credit counts; UNKNOWN scores zero unless gold is UNKNOWN.

  • HTML 75.9
  • Schema 50.4
  • AIPM 51.9

Metadata

Language, page kind, and freshness from page signals.

  • HTML 100
  • Schema 33.3
  • AIPM 66.7

Compression

Information delivered per token and context size. Higher means more matched answers for less context cost.

  • HTML 100
  • Schema 61.7
  • AIPM 60.6

Coverage map

AIPM matched 11 question(s); HTML matched 16. HTML/Schema (or a gap) still needed on Q6, Q8, Q10, Q11, Q12, Q13, Q14, Q15, Q19. No layer matched gold on Q13, Q14, Q15. Read this as complementary coverage — AIPM orients agents cheaply; HTML supplies depth when the sidecar cannot. HTML 16/20 matched (≈457 tok/match). Schema 11/20 matched (≈740 tok/match). AIPM 11/20 matched (≈753 tok/match). Figures are descriptive per layer — not a ranking. See per-question chains for sufficient layers, slice size, and measured cost.

AIPM matched

Q1, Q2, Q3, Q4, Q5, Q7, Q9, Q16, Q17, Q18, Q20

Alone: —

AIPM insufficient

Q6, Q8, Q10, Q11, Q12, Q13, Q14, Q15, Q19

HTML/Schema needed or all layers missed

Unanswered by all

Q13, Q14, Q15

HTML layer

Visible page text after stripping AIPM sidecars and JSON-LD. Measures what prose alone can answer.

Context fed: 7,126 chars

Tokens: 7,308 · median 1,899 ms

Matched this pack: 16/20

Stronger on

Understanding (85.7) · Retrieval (77.8) · Metadata (100) · Evidence (75.9) · Compression (100) · Answer Efficiency (100)

Weaker on

—

Schema layer

JSON-LD structured data with minimal page chrome. Measures what schema markup can answer.

Context fed: 10,757 chars

Tokens: 8,144 · median 2,128 ms

Matched this pack: 11/20

Stronger on

Understanding (85.7)

Weaker on

Metadata (33.3)

AIPM layer

AI Page Manifest (.ai.json) only. Measures what the machine layer can answer without HTML.

Context fed: 12,198 chars

Tokens: 8,287 · median 2,066 ms

Matched this pack: 11/20

Stronger on

Understanding (71.4)

Weaker on

—

When to use which layer

AIPM complements HTML — it does not replace full-page prose.

Scenario Recommended Why
Fast orientation (title, purpose, brand, intent) Compare machine card → HTML fallback on this run Machine-card pack: 8,287 tok · $0.00135. Routing sim saved ~6% tokens vs HTML-always.
Deep content / research (prose facts, process detail) HTML (with optional machine orientation) HTML pack: 7,308 tok · $0.00124. Use when depth needs body prose.
Structured entity pulls (org, location, typed fields) Schema.org Schema pack: 8,144 tok · $0.00132. Dense JSON-LD tends to score well here.

Findings

  • —Of 20 questions: lowest measured retrieval cost among sufficient layers — HTML 9 · Schema 3 · AIPM 5 · multi-layer 0 · unanswered 3. Descriptive counts only; no overall ranking.
  • —Semantic Redundancy 50% — the manifest repeats the same ideas across fields. This inflates structural size without helping per-question retrieval.
  • —Manifest design: Merge or differentiate `purpose` and `abstract` (100% overlap).
  • —Structural size is secondary. A larger AIPM is not a failure if per-question retrieval stays tiny — check Semantic Redundancy instead.
  • —AIPM matched 11 question(s); HTML matched 16. HTML/Schema (or a gap) still needed on Q6, Q8, Q10, Q11, Q12, Q13, Q14, Q15, Q19. No layer matched gold on Q13, Q14, Q15. Read this as complementary coverage — AIPM orients agents cheaply; HTML supplies depth when the sidecar cannot. HTML 16/20 matched (≈457 tok/match). Schema 11/20 matched (≈740 tok/match). AIPM 11/20 matched (≈753 tok/match). Figures are descriptive per layer — not a ranking. See per-question chains for sufficient layers, slice size, and measured cost.

Question-by-question layer analysis

Which layer(s) could answer; which need more or different context; minimum context fed this run.

Q1 · Understanding · orientation

What is this page or business?

Gold: Bilkent Kanal Açma

Gold source: html.h1|title|jsonld.name · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 131 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML title+h1 131 api 1 high yes
Schema name+headline+description 175 api 1 high yes
AIPM title+primaryTopic+purpose+abstract+publisher+keyFacts+audience+contentIntent 420 api 2 high yes
Execution chain · AIPM · 2 round(s)
  1. Question: What is this page or business?
  2. Planner: Planner mapped question `entity_what` on AIPM to minimal fields: title, primaryTopic.
  3. Slice: title+primaryTopic+purpose+abstract+publisher+keyFacts+audience+contentIntent — AIPM.title: Bilkent Kanal Açma AIPM.primaryTopic: Çankaya AIPM.purpose: Bilkent Kanal Açma hizmetl...
  4. Actual input tokens: 420
  5. LLM answer: This page is about Bilkent Kanal Açma services offered by FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET, focusing on cleaning channels without compromising quality.
  6. Evidence: AIPM.title: Bilkent Kanal Açma AIPM.primaryTopic: Çankaya AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden...
  7. Confidence: high · matched yes

Round 1: 115 in-tok · score 0 · miss

Round 2: 305 in-tok · score 88 · match

Execution chain · HTML · 1 round(s)
  1. Question: What is this page or business?
  2. Planner: Planner mapped question `entity_what` on HTML to minimal fields: title, h1.
  3. Slice: title+h1 — HTML.title: Bilkent Kanal Açma - Özdemir Tesisat 0 544 635 10 92 HTML.h1: Bilkent Kanal Açma
  4. Actual input tokens: 131
  5. LLM answer: This page is about Bilkent Kanal Açma, a service provided by Özdemir Tesisat.
  6. Evidence: HTML.title: Bilkent Kanal Açma - Özdemir Tesisat 0 544 635 10 92 HTML.h1: Bilkent Kanal Açma
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 1 round(s)
  1. Question: What is this page or business?
  2. Planner: Planner mapped question `entity_what` on SCHEMA to minimal fields: name, headline, description.
  3. Slice: name+headline+description — SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.headline: Bilkent Kanal Açma SCHEMA.des...
  4. Actual input tokens: 175
  5. LLM answer: This page is for a plumbing service business in Ankara called FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET, specializing in Bilkent Kanal Açma.
  6. Evidence: SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.headline: Bilkent Kanal Açma SCHEMA.description: Ankara'da, en iyi Bilkent Kana...
  7. Confidence: high · matched yes

HTML

sufficient

This page is about Bilkent Kanal Açma, a service provided by Özdemir Tesisat.

score 88 · 7,126 chars context · 131 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

This page is for a plumbing service business in Ankara called FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET, specializing in Bilkent Kanal Açma.

score 88 · 10,757 chars context · 175 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

This page is about Bilkent Kanal Açma services offered by FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET, focusing on cleaning channels without compromising quality.

score 88 · 12,198 chars context · 420 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q2 · Understanding · orientation

What is the main service or product offered?

Gold: Bilkent Kanal Açma

Gold source: html.h1|meta|prose · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost AIPM · 150 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML h1+title+meta_description+lead_paragraphs+prose_span+body_excerpt 407 api 2 high yes
Schema name+headline+description 177 api 1 high yes
AIPM primaryTopic+title+purpose 150 api 1 medium yes
Execution chain · AIPM · 1 round(s)
  1. Question: What is the main service or product offered?
  2. Planner: Planner mapped question `main_service` on AIPM to minimal fields: primaryTopic, title, purpose.
  3. Slice: primaryTopic+title+purpose — AIPM.primaryTopic: Çankaya AIPM.title: Bilkent Kanal Açma AIPM.purpose: Bilkent Kanal Açma hizmetl...
  4. Actual input tokens: 150
  5. LLM answer: The main service offered is channel cleaning.
  6. Evidence: AIPM.primaryTopic: Çankaya AIPM.title: Bilkent Kanal Açma AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden...
  7. Confidence: medium · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: What is the main service or product offered?
  2. Planner: Planner mapped question `main_service` on HTML to minimal fields: h1, title.
  3. Slice: h1+title+meta_description+lead_paragraphs+prose_span+body_excerpt — HTML.h1: Bilkent Kanal Açma HTML.title: Bilkent Kanal Açma - Özdemir Tesisat 0 544 635 10 92 HTML....
  4. Actual input tokens: 407
  5. LLM answer: Bilkent Kanal Açma hizmeti.
  6. Evidence: Bilkent Kanal Açma
  7. Confidence: high · matched yes

Round 1: 133 in-tok · score 0 · miss

Round 2: 274 in-tok · score 88 · match

Execution chain · SCHEMA · 1 round(s)
  1. Question: What is the main service or product offered?
  2. Planner: Planner mapped question `main_service` on SCHEMA to minimal fields: name, headline, description.
  3. Slice: name+headline+description — SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.headline: Bilkent Kanal Açma SCHEMA.des...
  4. Actual input tokens: 177
  5. LLM answer: Bilkent Kanal Açma / su tesisatçısı hizmeti.
  6. Evidence: Bilkent Kanal Açma
  7. Confidence: high · matched yes

HTML

sufficient

Bilkent Kanal Açma hizmeti.

score 88 · 7,126 chars context · 407 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

Bilkent Kanal Açma / su tesisatçısı hizmeti.

score 88 · 10,757 chars context · 177 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

The main service offered is channel cleaning.

score 72 · 12,198 chars context · 150 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q3 · Understanding · orientation

Who is the publisher or brand?

Gold: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET

Gold source: html.jsonld|title · html_independent

Sufficient: Schema, AIPM · Needs more / other context: HTML

Measured retrieval cost (API tokens) · lowest cost AIPM · 115 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML title+h1+meta_description+lead_paragraphs+prose_span+body_excerpt 402 api 2 low no
Schema publisher+name 135 api 1 high yes
AIPM publisher 115 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: Who is the publisher or brand?
  2. Planner: Planner mapped question `publisher` on AIPM to minimal fields: publisher.
  3. Slice: publisher — AIPM.publisher: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET
  4. Actual input tokens: 115
  5. LLM answer: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET.
  6. Evidence: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET
  7. Confidence: high · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: Who is the publisher or brand?
  2. Planner: Planner mapped question `publisher` on HTML to minimal fields: title.
  3. Slice: title+h1+meta_description+lead_paragraphs+prose_span+body_excerpt — HTML.title: Bilkent Kanal Açma - Özdemir Tesisat 0 544 635 10 92 HTML.h1: Bilkent Kanal Açma HTML....
  4. Actual input tokens: 402
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 120 in-tok · score 0 · miss

Round 2: 282 in-tok · score 0 · miss

Execution chain · SCHEMA · 1 round(s)
  1. Question: Who is the publisher or brand?
  2. Planner: Planner mapped question `publisher` on SCHEMA to minimal fields: publisher, name.
  3. Slice: publisher+name — SCHEMA.publisher: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YA...
  4. Actual input tokens: 135
  5. LLM answer: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET.
  6. Evidence: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET
  7. Confidence: high · matched yes

HTML

insufficient

UNKNOWN.

score 0 · 7,126 chars context · 402 in-tokens

HTML did not answer from 7126 chars of context — additional or different layer context needed.

Schema

sufficient

FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET.

score 100 · 10,757 chars context · 135 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET.

score 100 · 12,198 chars context · 115 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q4 · Retrieval · depth

Where does this business operate or which city is mentioned?

Gold: Ankara

Gold source: html.address|prose|jsonld · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost SCHEMA · 107 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML prose_span+lead_paragraphs 141 api 1 high yes
Schema addressLocality 107 api 1 high yes
AIPM purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent 320 api 2 high yes
Execution chain · AIPM · 2 round(s)
  1. Question: Where does this business operate or which city is mentioned?
  2. Planner: Planner mapped question `geo_signal` on AIPM to minimal fields: geo, areaServed, address.
  3. Slice: purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent — AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizl...
  4. Actual input tokens: 320
  5. LLM answer: Çankaya.
  6. Evidence: AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizletebilirsiniz. AIPM.abstract: Bilkent K...
  7. Confidence: high · matched yes

Round 1: 0 in-tok · score 0 · miss

Round 2: 320 in-tok · score 100 · match

Execution chain · HTML · 1 round(s)
  1. Question: Where does this business operate or which city is mentioned?
  2. Planner: Planner mapped question `geo_signal` on HTML to minimal fields: prose_span, lead_paragraphs.
  3. Slice: prose_span+lead_paragraphs — HTML.prose_span: Ankara HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUV...
  4. Actual input tokens: 141
  5. LLM answer: Ankara
  6. Evidence: Ankara
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 1 round(s)
  1. Question: Where does this business operate or which city is mentioned?
  2. Planner: Planner mapped question `geo_signal` on SCHEMA to minimal fields: address, addressLocality, areaServed.
  3. Slice: addressLocality — SCHEMA.addressLocality: Etimesgut
  4. Actual input tokens: 107
  5. LLM answer: Etimesgut
  6. Evidence: Etimesgut
  7. Confidence: high · matched yes

HTML

sufficient

Ankara

score 100 · 7,126 chars context · 141 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

Etimesgut

score 100 · 10,757 chars context · 107 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

Çankaya.

score 100 · 12,198 chars context · 320 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q5 · Understanding · orientation

Who is the intended audience of this page?

Gold: customers seeking this service

Gold source: html.prose|meta · html_independent

Sufficient: HTML, AIPM · Needs more / other context: Schema

Measured retrieval cost (API tokens) · lowest cost AIPM · 189 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt 452 api 2 high yes
Schema description+name+headline+publisher+@type+block 484 api 2 low no
AIPM audience+purpose 189 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: Who is the intended audience of this page?
  2. Planner: Planner mapped question `audience` on AIPM to minimal fields: audience, purpose.
  3. Slice: audience+purpose — AIPM.audience: ["sıhhi tesisat alanında","kanal açma","klozet tamiri","petek temizleme","tesisat yen...
  4. Actual input tokens: 189
  5. LLM answer: The intended audience of this page includes individuals and institutions in need of plumbing services such as channel opening, toilet repair, radiator cleaning,...
  6. Evidence: AIPM.audience: ["sıhhi tesisat alanında","kanal açma","klozet tamiri","petek temizleme","tesisat yenileme","su kaçak tespiti hizmetlerine 7/...
  7. Confidence: high · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: Who is the intended audience of this page?
  2. Planner: Planner mapped question `audience` on HTML to minimal fields: meta_description, lead_paragraphs.
  3. Slice: meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt — HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI ol...
  4. Actual input tokens: 452
  5. LLM answer: customers seeking this service
  6. Evidence: customers seeking this service
  7. Confidence: high · matched yes

Round 1: 180 in-tok · score 0 · miss

Round 2: 272 in-tok · score 100 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: Who is the intended audience of this page?
  2. Planner: Planner mapped question `audience` on SCHEMA to minimal fields: description, name.
  3. Slice: description+name+headline+publisher+@type+block — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 484
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 165 in-tok · score 0 · miss

Round 2: 319 in-tok · score 0 · miss

HTML

sufficient

customers seeking this service

score 100 · 7,126 chars context · 452 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 484 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

sufficient

The intended audience of this page includes individuals and institutions in need of plumbing services such as channel opening, toilet repair, radiator cleaning, plumbing renewal, and leak detection services available 24/7.

score 100 · 12,198 chars context · 189 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q6 · Understanding · orientation

What is the purpose of this page?

Gold: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmaktayız.

Gold source: html.page_summary · html_independent

Sufficient: HTML, Schema · Needs more / other context: AIPM

Measured retrieval cost (API tokens) · lowest cost SCHEMA · 164 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt 490 api 2 high yes
Schema description+name 164 api 1 high yes
AIPM purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent 478 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What is the purpose of this page?
  2. Planner: Planner mapped question `purpose` on AIPM to minimal fields: purpose.
  3. Slice: purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent — AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizl...
  4. Actual input tokens: 478
  5. LLM answer: The purpose of this page is to offer quality canal cleaning services without compromising on standards.
  6. Evidence: AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizletebilirsiniz. AIPM.abstract: Bilkent K...
  7. Confidence: low · matched no

Round 1: 127 in-tok · score 20 · miss

Round 2: 351 in-tok · score 20 · miss

Execution chain · HTML · 2 round(s)
  1. Question: What is the purpose of this page?
  2. Planner: Planner mapped question `purpose` on HTML to minimal fields: meta_description, lead_paragraphs.
  3. Slice: meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt — HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI ol...
  4. Actual input tokens: 490
  5. LLM answer: The purpose of this page is to offer effective plumbing services, specifically for Bilkent Kanal Açma and related plumbing tasks in Ankara.
  6. Evidence: HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmak...
  7. Confidence: high · matched yes

Round 1: 179 in-tok · score 0 · miss

Round 2: 311 in-tok · score 100 · match

Execution chain · SCHEMA · 1 round(s)
  1. Question: What is the purpose of this page?
  2. Planner: Planner mapped question `purpose` on SCHEMA to minimal fields: description, name.
  3. Slice: description+name — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 164
  5. LLM answer: The purpose of this page is to provide effective plumbing services in Ankara as a top Bilkent canal opening and equipped plumber.
  6. Evidence: SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmaktay...
  7. Confidence: high · matched yes

Root cause · Layer miss

Improve orientation fields; do not paste full HTML into the sidecar.

HTML

sufficient

The purpose of this page is to offer effective plumbing services, specifically for Bilkent Kanal Açma and related plumbing tasks in Ankara.

score 100 · 7,126 chars context · 490 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

The purpose of this page is to provide effective plumbing services in Ankara as a top Bilkent canal opening and equipped plumber.

score 80 · 10,757 chars context · 164 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

insufficient

The purpose of this page is to offer quality canal cleaning services without compromising on standards.

score 20 · 12,198 chars context · 478 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q7 · Understanding · orientation

Summarize this page in one sentence.

Gold: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmaktayız.

Gold source: html.page_summary · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost AIPM · 160 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs 492 api 2 medium yes
Schema description+name 165 api 1 medium yes
AIPM abstract+purpose 160 api 1 medium yes
Execution chain · AIPM · 1 round(s)
  1. Question: Summarize this page in one sentence.
  2. Planner: Planner mapped question `abstract` on AIPM to minimal fields: abstract, purpose.
  3. Slice: abstract+purpose — AIPM.abstract: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 160
  5. LLM answer: Bilkent Kanal Açma offers quality services for cleaning channels without compromising on standards.
  6. Evidence: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizletebilirsiniz.
  7. Confidence: medium · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: Summarize this page in one sentence.
  2. Planner: Planner mapped question `abstract` on HTML to minimal fields: meta_description, lead_paragraphs.
  3. Slice: meta_description+lead_paragraphs — HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI ol...
  4. Actual input tokens: 492
  5. LLM answer: Ankara'da etkili sıhhi tesisat hizmetleri sunan bir firma, Bilkent Kanal Açma ve cihazlı tesisat hizmetleri ile Etlik Su Tesisatçısı olarak hizmet vermektedir.
  6. Evidence: HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmak...
  7. Confidence: medium · matched yes

Round 1: 180 in-tok · score 55 · match

Round 2: 312 in-tok · score 52 · match

Execution chain · SCHEMA · 1 round(s)
  1. Question: Summarize this page in one sentence.
  2. Planner: Planner mapped question `abstract` on SCHEMA to minimal fields: description, name.
  3. Slice: description+name — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 165
  5. LLM answer: Ankara'da etkili sıhhi tesisat hizmeti sunan FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET, en iyi Bilkent Kanal Açma ve cihazlı tesisatçı olarak hizmet vermektedir.
  6. Evidence: SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmaktay...
  7. Confidence: medium · matched yes

HTML

sufficient

Ankara'da etkili sıhhi tesisat hizmetleri sunan bir firma, Bilkent Kanal Açma ve cihazlı tesisat hizmetleri ile Etlik Su Tesisatçısı olarak hizmet vermektedir.

score 55 · 7,126 chars context · 492 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

Ankara'da etkili sıhhi tesisat hizmeti sunan FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET, en iyi Bilkent Kanal Açma ve cihazlı tesisatçı olarak hizmet vermektedir.

score 65 · 10,757 chars context · 165 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

Bilkent Kanal Açma offers quality services for cleaning channels without compromising on standards.

score 70 · 12,198 chars context · 160 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q8 · Understanding · orientation

What is the primary topic or headline of this page?

Gold: Bilkent Kanal Açma

Gold source: html.h1|title · html_independent

Sufficient: HTML, Schema · Needs more / other context: AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 135 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML h1+title 135 api 1 high yes
Schema name+headline+description 179 api 1 high yes
AIPM primaryTopic+title+purpose+abstract+publisher+keyFacts+audience+contentIntent 461 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What is the primary topic or headline of this page?
  2. Planner: Planner mapped question `primary_topic` on AIPM to minimal fields: primaryTopic, title, purpose.
  3. Slice: primaryTopic+title+purpose+abstract+publisher+keyFacts+audience+contentIntent — AIPM.primaryTopic: Çankaya AIPM.title: Bilkent Kanal Açma AIPM.purpose: Bilkent Kanal Açma hizmetl...
  4. Actual input tokens: 461
  5. LLM answer: Çankaya
  6. Evidence: Çankaya
  7. Confidence: low · matched no

Round 1: 152 in-tok · score 0 · miss

Round 2: 309 in-tok · score 0 · miss

Execution chain · HTML · 1 round(s)
  1. Question: What is the primary topic or headline of this page?
  2. Planner: Planner mapped question `primary_topic` on HTML to minimal fields: h1, title.
  3. Slice: h1+title — HTML.h1: Bilkent Kanal Açma HTML.title: Bilkent Kanal Açma - Özdemir Tesisat 0 544 635 10 92
  4. Actual input tokens: 135
  5. LLM answer: Bilkent Kanal Açma
  6. Evidence: Bilkent Kanal Açma
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 1 round(s)
  1. Question: What is the primary topic or headline of this page?
  2. Planner: Planner mapped question `primary_topic` on SCHEMA to minimal fields: name, headline, description.
  3. Slice: name+headline+description — SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.headline: Bilkent Kanal Açma SCHEMA.des...
  4. Actual input tokens: 179
  5. LLM answer: Bilkent Kanal Açma
  6. Evidence: Bilkent Kanal Açma
  7. Confidence: high · matched yes

Root cause · Layer miss

Improve orientation fields; do not paste full HTML into the sidecar.

HTML

sufficient

Bilkent Kanal Açma

score 100 · 7,126 chars context · 135 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

Bilkent Kanal Açma

score 100 · 10,757 chars context · 179 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

insufficient

Çankaya

score 0 · 12,198 chars context · 461 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q9 · Metadata · orientation

What language is this page written in?

Gold: tr

Gold source: html.lang · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 101 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lang 101 api 1 high yes
Schema inLanguage 103 api 1 high yes
AIPM inLanguage 102 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: What language is this page written in?
  2. Planner: Planner mapped question `language` on AIPM to minimal fields: inLanguage.
  3. Slice: inLanguage — AIPM.inLanguage: tr
  4. Actual input tokens: 102
  5. LLM answer: The page is written in Turkish (tr).
  6. Evidence: AIPM.inLanguage: tr
  7. Confidence: high · matched yes
Execution chain · HTML · 1 round(s)
  1. Question: What language is this page written in?
  2. Planner: Planner mapped question `language` on HTML to minimal fields: lang.
  3. Slice: lang — HTML.lang: lang=tr
  4. Actual input tokens: 101
  5. LLM answer: The page is written in Turkish.
  6. Evidence: HTML.lang: lang=tr
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 1 round(s)
  1. Question: What language is this page written in?
  2. Planner: Planner mapped question `language` on SCHEMA to minimal fields: inLanguage.
  3. Slice: inLanguage — SCHEMA.inLanguage: tr-TR
  4. Actual input tokens: 103
  5. LLM answer: The page is written in Turkish (tr-TR).
  6. Evidence: SCHEMA.inLanguage: tr-TR
  7. Confidence: high · matched yes

HTML

sufficient

The page is written in Turkish.

score 88 · 7,126 chars context · 101 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

The page is written in Turkish (tr-TR).

score 88 · 10,757 chars context · 103 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

The page is written in Turkish (tr).

score 88 · 12,198 chars context · 102 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q10 · Retrieval · depth

What phone number or contact detail is listed?

Gold: +905446351092

Gold source: html.title|prose|jsonld.telephone · html_independent

Sufficient: HTML, Schema · Needs more / other context: AIPM

Measured retrieval cost (API tokens) · lowest cost SCHEMA · 329 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML title+lead_paragraphs+h1+meta_description+prose_span+body_excerpt 431 api 2 high yes
Schema telephone+block+description 329 api 1 high yes
AIPM keyFacts+publisher+abstract+purpose+primaryTopic+title+audience+contentIntent 579 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What phone number or contact detail is listed?
  2. Planner: Planner mapped question `contact_phone` on AIPM to minimal fields: keyFacts, publisher, abstract.
  3. Slice: keyFacts+publisher+abstract+purpose+primaryTopic+title+audience+contentIntent — AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 579
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 227 in-tok · score 0 · miss

Round 2: 352 in-tok · score 0 · miss

Execution chain · HTML · 2 round(s)
  1. Question: What phone number or contact detail is listed?
  2. Planner: Planner mapped question `contact_phone` on HTML to minimal fields: title, lead_paragraphs.
  3. Slice: title+lead_paragraphs+h1+meta_description+prose_span+body_excerpt — HTML.title: Bilkent Kanal Açma - Özdemir Tesisat 0 544 635 10 92 HTML.lead_paragraphs: Etlik Su Tes...
  4. Actual input tokens: 431
  5. LLM answer: +905446351092
  6. Evidence: +905446351092
  7. Confidence: high · matched yes

Round 1: 158 in-tok · score 0 · miss

Round 2: 273 in-tok · score 100 · match

Execution chain · SCHEMA · 1 round(s)
  1. Question: What phone number or contact detail is listed?
  2. Planner: Planner mapped question `contact_phone` on SCHEMA to minimal fields: telephone, block, description.
  3. Slice: telephone+block+description — SCHEMA.telephone: +905446351092 SCHEMA.block: { "@context":"https://schema.org", "@graph":[...
  4. Actual input tokens: 329
  5. LLM answer: +905446351092
  6. Evidence: +905446351092
  7. Confidence: high · matched yes

Root cause · Depth gap (expected)

Prose/detail questions belong to HTML — escalate after AIPM orientation.

HTML

sufficient

+905446351092

score 100 · 7,126 chars context · 431 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

+905446351092

score 100 · 10,757 chars context · 329 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 579 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q11 · Retrieval · depth

What physical address is listed for this business?

Gold: Eryaman Life Ap, Elvan, Ahi Elvan Cd. No:20/A

Gold source: html.prose|jsonld.address · html_independent

Sufficient: HTML · Needs more / other context: Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 156 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML prose_span+lead_paragraphs 156 api 1 high yes
Schema addressLocality+name+headline+description+publisher+@type+block 494 api 2 low no
AIPM purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent 352 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What physical address is listed for this business?
  2. Planner: Planner mapped question `address` on AIPM to minimal fields: address, geo, areaServed.
  3. Slice: purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent — AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizl...
  4. Actual input tokens: 352
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 0 in-tok · score 0 · miss

Round 2: 352 in-tok · score 0 · miss

Execution chain · HTML · 1 round(s)
  1. Question: What physical address is listed for this business?
  2. Planner: Planner mapped question `address` on HTML to minimal fields: prose_span, lead_paragraphs.
  3. Slice: prose_span+lead_paragraphs — HTML.prose_span: Eryaman Life Ap, Elvan, Ahi Elvan Cd. No:20/A HTML.lead_paragraphs: Etlik Su Tesis...
  4. Actual input tokens: 156
  5. LLM answer: Eryaman Life Ap, Elvan, Ahi Elvan Cd. No:20/A
  6. Evidence: Eryaman Life Ap, Elvan, Ahi Elvan Cd. No:20/A
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: What physical address is listed for this business?
  2. Planner: Planner mapped question `address` on SCHEMA to minimal fields: address, addressLocality, areaServed.
  3. Slice: addressLocality+name+headline+description+publisher+@type+block — SCHEMA.addressLocality: Etimesgut SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.headl...
  4. Actual input tokens: 494
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 105 in-tok · score 0 · miss

Round 2: 389 in-tok · score 0 · miss

Root cause · Depth gap (expected)

Prose/detail questions belong to HTML — escalate after AIPM orientation.

HTML

sufficient

Eryaman Life Ap, Elvan, Ahi Elvan Cd. No:20/A

score 100 · 7,126 chars context · 156 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 494 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 352 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q12 · Retrieval · depth

What are the opening hours or availability?

Gold: 24/7

Gold source: html.prose|jsonld.openingHours · html_independent

Sufficient: HTML · Needs more / other context: Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 183 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span 183 api 1 high yes
Schema block+description+name+headline+publisher+@type 575 api 2 low no
AIPM keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent 588 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What are the opening hours or availability?
  2. Planner: Planner mapped question `hours` on AIPM to minimal fields: keyFacts, abstract, purpose.
  3. Slice: keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 588
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 237 in-tok · score 0 · miss

Round 2: 351 in-tok · score 0 · miss

Execution chain · HTML · 1 round(s)
  1. Question: What are the opening hours or availability?
  2. Planner: Planner mapped question `hours` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 183
  5. LLM answer: 24/7
  6. Evidence: 24/7
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: What are the opening hours or availability?
  2. Planner: Planner mapped question `hours` on SCHEMA to minimal fields: openingHours, block, description.
  3. Slice: block+description+name+headline+publisher+@type — SCHEMA.block: { "@context":"https://schema.org", "@graph":[ { "@type":"Organization",...
  4. Actual input tokens: 575
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 257 in-tok · score 0 · miss

Round 2: 318 in-tok · score 0 · miss

Root cause · Depth gap (expected)

Prose/detail questions belong to HTML — escalate after AIPM orientation.

HTML

sufficient

24/7

score 100 · 7,126 chars context · 183 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 575 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 588 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q13 · Retrieval · depth

Is pricing or package information mentioned, and what is it?

Gold: .woocommerce ul.products li.product .price, .woocommerce ul.products li.product .woocommerce-loop-prod

Gold source: html.prose|jsonld.offers · html_independent

No layer provided a sufficient answer from its context alone.

Measured retrieval cost (API tokens) · lowest cost HTML · 493 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description 493 api 2 low no
Schema block+description+name+headline+publisher+@type 583 api 2 low no
AIPM keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent 563 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: Is pricing or package information mentioned, and what is it?
  2. Planner: Planner mapped question `pricing` on AIPM to minimal fields: keyFacts, sections, abstract.
  3. Slice: keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 563
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 208 in-tok · score 0 · miss

Round 2: 355 in-tok · score 0 · miss

Execution chain · HTML · 2 round(s)
  1. Question: Is pricing or package information mentioned, and what is it?
  2. Planner: Planner mapped question `pricing` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 493
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 203 in-tok · score 0 · miss

Round 2: 290 in-tok · score 0 · miss

Execution chain · SCHEMA · 2 round(s)
  1. Question: Is pricing or package information mentioned, and what is it?
  2. Planner: Planner mapped question `pricing` on SCHEMA to minimal fields: offers, price, block, description.
  3. Slice: block+description+name+headline+publisher+@type — SCHEMA.block: { "@context":"https://schema.org", "@graph":[ { "@type":"Organization",...
  4. Actual input tokens: 583
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 261 in-tok · score 0 · miss

Round 2: 322 in-tok · score 0 · miss

Root cause · Depth gap (expected)

Prose/detail questions belong to HTML — escalate after AIPM orientation.

HTML

insufficient

UNKNOWN.

score 0 · 7,126 chars context · 493 in-tokens

HTML did not answer from 7126 chars of context — additional or different layer context needed.

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 583 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 563 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q14 · Retrieval · depth

What makes this offer different or stand out?

Gold: son teknoloji

Gold source: html.prose|meta · html_independent

No layer provided a sufficient answer from its context alone.

Measured retrieval cost (API tokens) · lowest cost HTML · 453 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description 453 api 2 low no
Schema description+block+name+headline+publisher+@type 598 api 2 low no
AIPM keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent 590 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What makes this offer different or stand out?
  2. Planner: Planner mapped question `differentiator` on AIPM to minimal fields: keyFacts, abstract, purpose.
  3. Slice: keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 590
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 238 in-tok · score 0 · miss

Round 2: 352 in-tok · score 0 · miss

Execution chain · HTML · 2 round(s)
  1. Question: What makes this offer different or stand out?
  2. Planner: Planner mapped question `differentiator` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 453
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 183 in-tok · score 0 · miss

Round 2: 270 in-tok · score 0 · miss

Execution chain · SCHEMA · 2 round(s)
  1. Question: What makes this offer different or stand out?
  2. Planner: Planner mapped question `differentiator` on SCHEMA to minimal fields: description, block, name.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 598
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 279 in-tok · score 0 · miss

Round 2: 319 in-tok · score 0 · miss

Root cause · Depth gap (expected)

Prose/detail questions belong to HTML — escalate after AIPM orientation.

HTML

insufficient

UNKNOWN.

score 0 · 7,126 chars context · 453 in-tokens

HTML did not answer from 7126 chars of context — additional or different layer context needed.

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 598 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 590 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q15 · Evidence · depth

What trust signals are present (reviews, certificates, guarantees, years of experience)?

Gold: review

Gold source: html.prose|jsonld · html_independent

No layer provided a sufficient answer from its context alone.

Measured retrieval cost (API tokens) · lowest cost HTML · 465 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description 465 api 2 low no
Schema description+block+name+headline+publisher+@type 732 api 2 low no
AIPM keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent 604 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What trust signals are present (reviews, certificates, guarantees, years of experience)?
  2. Planner: Planner mapped question `trust` on AIPM to minimal fields: keyFacts, abstract, purpose.
  3. Slice: keyFacts+abstract+purpose+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 604
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 245 in-tok · score 0 · miss

Round 2: 359 in-tok · score 0 · miss

Execution chain · HTML · 2 round(s)
  1. Question: What trust signals are present (reviews, certificates, guarantees, years of experience)?
  2. Planner: Planner mapped question `trust` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 465
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 189 in-tok · score 0 · miss

Round 2: 276 in-tok · score 0 · miss

Execution chain · SCHEMA · 2 round(s)
  1. Question: What trust signals are present (reviews, certificates, guarantees, years of experience)?
  2. Planner: Planner mapped question `trust` on SCHEMA to minimal fields: description, block, name.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 732
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 346 in-tok · score 0 · miss

Round 2: 386 in-tok · score 0 · miss

Root cause · Depth gap (expected)

Prose/detail questions belong to HTML — escalate after AIPM orientation.

HTML

insufficient

UNKNOWN.

score 0 · 7,126 chars context · 465 in-tokens

HTML did not answer from 7126 chars of context — additional or different layer context needed.

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 732 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 604 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q16 · Retrieval · depth

What frequently asked questions or FAQ topics are covered?

Gold: UNKNOWN

Gold source: html.faq|jsonld.FAQPage · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 453 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description 453 api 2 low yes
Schema block+description+name+headline+publisher+@type 579 api 2 low yes
AIPM faqEntity+keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent 543 api 2 low yes
Execution chain · AIPM · 2 round(s)
  1. Question: What frequently asked questions or FAQ topics are covered?
  2. Planner: Planner mapped question `faq` on AIPM to minimal fields: faqEntity, keyFacts.
  3. Slice: faqEntity+keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent — AIPM.faqEntity: [] AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden...
  4. Actual input tokens: 543
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched yes

Round 1: 182 in-tok · score 100 · match

Round 2: 361 in-tok · score 100 · match

Execution chain · HTML · 2 round(s)
  1. Question: What frequently asked questions or FAQ topics are covered?
  2. Planner: Planner mapped question `faq` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 453
  5. LLM answer: UNKNOWN
  6. Evidence: —
  7. Confidence: low · matched yes

Round 1: 183 in-tok · score 100 · match

Round 2: 270 in-tok · score 100 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: What frequently asked questions or FAQ topics are covered?
  2. Planner: Planner mapped question `faq` on SCHEMA to minimal fields: block, description.
  3. Slice: block+description+name+headline+publisher+@type — SCHEMA.block: { "@context":"https://schema.org", "@graph":[ { "@type":"Organization",...
  4. Actual input tokens: 579
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched yes

Round 1: 259 in-tok · score 100 · match

Round 2: 320 in-tok · score 100 · match

HTML

sufficient

UNKNOWN

score 100 · 7,126 chars context · 453 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

UNKNOWN.

score 100 · 10,757 chars context · 579 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

UNKNOWN.

score 100 · 12,198 chars context · 543 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q17 · Retrieval · depth

What action should the visitor take (call, buy, contact, book)?

Gold: UNKNOWN

Gold source: html.cta|prose · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 463 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description 463 api 2 low yes
Schema description+block+name+headline+publisher+@type 610 api 2 low yes
AIPM purpose+abstract+keyFacts+primaryTopic+title+publisher+audience+contentIntent 602 api 2 low yes
Execution chain · AIPM · 2 round(s)
  1. Question: What action should the visitor take (call, buy, contact, book)?
  2. Planner: Planner mapped question `cta` on AIPM to minimal fields: purpose, abstract, keyFacts.
  3. Slice: purpose+abstract+keyFacts+primaryTopic+title+publisher+audience+contentIntent — AIPM.purpose: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizl...
  4. Actual input tokens: 602
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched yes

Round 1: 244 in-tok · score 100 · match

Round 2: 358 in-tok · score 100 · match

Execution chain · HTML · 2 round(s)
  1. Question: What action should the visitor take (call, buy, contact, book)?
  2. Planner: Planner mapped question `cta` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 463
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched yes

Round 1: 188 in-tok · score 100 · match

Round 2: 275 in-tok · score 100 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: What action should the visitor take (call, buy, contact, book)?
  2. Planner: Planner mapped question `cta` on SCHEMA to minimal fields: description, block, name.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 610
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched yes

Round 1: 285 in-tok · score 100 · match

Round 2: 325 in-tok · score 100 · match

HTML

sufficient

UNKNOWN.

score 100 · 7,126 chars context · 463 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

sufficient

UNKNOWN.

score 100 · 10,757 chars context · 610 in-tokens

Schema answered using 10757 chars of layer context (minimum fed this run).

AIPM

sufficient

UNKNOWN.

score 100 · 12,198 chars context · 602 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q18 · Retrieval · depth

What related products, services, or pages are mentioned?

Gold: Cancel Preloader, Su Kaçak Tespiti, Kanal Açma

Gold source: html.nav|prose · html_independent

Sufficient: HTML, AIPM · Needs more / other context: Schema

Measured retrieval cost (API tokens) · lowest cost HTML · 198 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span 198 api 1 high yes
Schema description+block+name+headline+publisher+@type 602 api 2 low no
AIPM keyFacts+primaryTopic+purpose+abstract+title+publisher+audience+contentIntent 539 api 2 high yes
Execution chain · AIPM · 2 round(s)
  1. Question: What related products, services, or pages are mentioned?
  2. Planner: Planner mapped question `related` on AIPM to minimal fields: sections, keyFacts, primaryTopic.
  3. Slice: keyFacts+primaryTopic+purpose+abstract+title+publisher+audience+contentIntent — AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 539
  5. LLM answer: The related services mentioned are kanal açma, klozet tamiri, petek temizleme, tesisat yenileme, and su kaçak tespiti.
  6. Evidence: AIPM.keyFacts: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temizletebilirsiniz. This blog post is publis...
  7. Confidence: high · matched yes

Round 1: 185 in-tok · score 0 · miss

Round 2: 354 in-tok · score 100 · match

Execution chain · HTML · 1 round(s)
  1. Question: What related products, services, or pages are mentioned?
  2. Planner: Planner mapped question `related` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 198
  5. LLM answer: Etlik Su Tesisatçısı, Alaturka tuvaleti alafranga tuvalete çevirme, klozet tıkanıklığı açma, su kaçak tespiti, kanal açma.
  6. Evidence: Etlik Su Tesisatçısı
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: What related products, services, or pages are mentioned?
  2. Planner: Planner mapped question `related` on SCHEMA to minimal fields: description, block, name.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olara...
  4. Actual input tokens: 602
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 281 in-tok · score 0 · miss

Round 2: 321 in-tok · score 0 · miss

HTML

sufficient

Etlik Su Tesisatçısı, Alaturka tuvaleti alafranga tuvalete çevirme, klozet tıkanıklığı açma, su kaçak tespiti, kanal açma.

score 100 · 7,126 chars context · 198 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 602 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

sufficient

The related services mentioned are kanal açma, klozet tamiri, petek temizleme, tesisat yenileme, and su kaçak tespiti.

score 100 · 12,198 chars context · 539 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Q19 · Metadata · orientation

When was this page published or last updated?

Gold: 2022-10-08T23:41:48+00:00

Gold source: html.time|jsonld.dateModified · html_independent

Sufficient: HTML · Needs more / other context: Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 481 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description 481 api 2 high yes
Schema block+description+name+headline+publisher+@type 697 api 2 low no
AIPM abstract+keyFacts+title+purpose+primaryTopic+publisher+audience+contentIntent 569 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: When was this page published or last updated?
  2. Planner: Planner mapped question `freshness` on AIPM to minimal fields: abstract, keyFacts, title.
  3. Slice: abstract+keyFacts+title+purpose+primaryTopic+publisher+audience+contentIntent — AIPM.abstract: Bilkent Kanal Açma hizmetlerimizde siz de kaliteden ödün vermeden kanallarınızı temiz...
  4. Actual input tokens: 569
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 217 in-tok · score 0 · miss

Round 2: 352 in-tok · score 0 · miss

Execution chain · HTML · 2 round(s)
  1. Question: When was this page published or last updated?
  2. Planner: Planner mapped question `freshness` on HTML to minimal fields: lead_paragraphs, body_excerpt, prose_span.
  3. Slice: lead_paragraphs+body_excerpt+prose_span+title+h1+meta_description — HTML.lead_paragraphs: Etlik Su Tesisatçısı ALATURKA TUVALETİ ALAFRANGA TUVALETE ÇEVİRME HTML.body_...
  4. Actual input tokens: 481
  5. LLM answer: 2022-10-08T23:41:48+00:00
  6. Evidence: 2022-10-08T23:41:48+00:00
  7. Confidence: high · matched yes

Round 1: 197 in-tok · score 31 · miss

Round 2: 284 in-tok · score 100 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: When was this page published or last updated?
  2. Planner: Planner mapped question `freshness` on SCHEMA to minimal fields: block, description.
  3. Slice: block+description+name+headline+publisher+@type — SCHEMA.block: { "@context":"https://schema.org", "@graph":[ { "@type":"Organization",...
  4. Actual input tokens: 697
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 318 in-tok · score 0 · miss

Round 2: 379 in-tok · score 0 · miss

Root cause · Missing in manifest

Add or fix the corresponding AIPM field (title, purpose, primaryTopic, publisher, inLanguage).

HTML

sufficient

2022-10-08T23:41:48+00:00

score 100 · 7,126 chars context · 481 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 697 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 12,198 chars context · 569 in-tokens

AIPM did not answer from 12198 chars of context — additional or different layer context needed.

Q20 · Metadata · orientation

What kind of page is this (article, product, local service, company page, FAQ, other)?

Gold: product

Gold source: html.heuristic|jsonld.@type · html_independent

Sufficient: HTML, AIPM · Needs more / other context: Schema

Measured retrieval cost (API tokens) · lowest cost AIPM · 124 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt 473 api 2 high yes
Schema @type+name+headline+description+publisher+block 445 api 2 low no
AIPM contentIntent+pageType 124 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: What kind of page is this (article, product, local service, company page, FAQ, other)?
  2. Planner: Planner mapped question `page_kind` on AIPM to minimal fields: contentIntent, pageType.
  3. Slice: contentIntent+pageType — AIPM.contentIntent: informational AIPM.pageType: blog
  4. Actual input tokens: 124
  5. LLM answer: blog
  6. Evidence: blog
  7. Confidence: high · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: What kind of page is this (article, product, local service, company page, FAQ, other)?
  2. Planner: Planner mapped question `page_kind` on HTML to minimal fields: meta_description, lead_paragraphs.
  3. Slice: meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt — HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI ol...
  4. Actual input tokens: 473
  5. LLM answer: local service
  6. Evidence: HTML.meta_description: Ankara'da, en iyi Bilkent Kanal Açma / su tesisatçısı ve CİHAZLI TESİSATÇI olarak etkili sıhhi tesisat hizmeti sunmak...
  7. Confidence: high · matched yes

Round 1: 192 in-tok · score 0 · miss

Round 2: 281 in-tok · score 100 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: What kind of page is this (article, product, local service, company page, FAQ, other)?
  2. Planner: Planner mapped question `page_kind` on SCHEMA to minimal fields: @type.
  3. Slice: @type+name+headline+description+publisher+block — SCHEMA.@type: Organization SCHEMA.name: FİLLİ BOYA ÖZDEMİR TESİSAT YAPI MARKET SCHEMA.headline: Bi...
  4. Actual input tokens: 445
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 114 in-tok · score 0 · miss

Round 2: 331 in-tok · score 0 · miss

HTML

sufficient

local service

score 100 · 7,126 chars context · 473 in-tokens

HTML answered using 7126 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 10,757 chars context · 445 in-tokens

Schema did not answer from 10757 chars of context — additional or different layer context needed.

AIPM

sufficient

blog

score 100 · 12,198 chars context · 124 in-tokens

AIPM answered using 12198 chars of layer context (minimum fed this run).

Methodology

  • AI Context Benchmark: Planner → Slice → LLM with measured API input tokens.
  • Reports describe sufficient layers, slice size, cost, and evidence — they do not declare a winning format.
  • Layers under test today: HTML, Schema.org, AIPM (extensible to RSS, Markdown, PDF, …).
  • Engine aipm_benchmark_v4 · Wed, Jul 29, 2026 3:01 PM · aipm_benchmark_score_v10

Share

Per-question context chain across layers.

Post on X LinkedIn WhatsApp
About the lab →