AigeoRadar

AigeoRadar · AI Context Benchmark

AI Context Benchmark

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

Free Full LLMS.txt Generator

https://aigeoradar.com/blog/free-full-llmstxt-generator-create-a-complete-ai-ready-llmstxt

12 questions Answer LLM: unknown ai_context_benchmark_v1

Of 12 questions: lowest measured retrieval cost among sufficient layers — HTML 4 · Schema 1 · AIPM 7 · multi-layer 0 · unanswered 0. Descriptive counts only; no overall ranking.

Of 12 questions: lowest measured retrieval cost among sufficient layers — HTML 4 · Schema 1 · AIPM 7 · multi-layer 0 · unanswered 0. 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 28% — moderate field overlap; trim duplicated purpose/summary/keyFacts wording.

Gold mode · Mixed independent + AIPM consistency

Most Understanding/Retrieval gold comes from HTML. Some questions remain AIPM self-consistency checks (tagged) and are excluded from Understanding when possible. Matching a machine card against its own fields is not evidence that that layer outperforms HTML.

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
unknown / unknown (single provider this lab)
Model snapshot
2026-07
Run id
5ed5ffc3-b428-4f12-a466-fee59c601b34

Question outcomes

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

4

Lowest cost: HTML

1

Lowest cost: Schema

7

Lowest cost: AIPM

0

Multi-layer

0

Unanswered

Manifest design

Semantic Redundancy · 28%

Semantic Redundancy 28% — moderate field overlap; trim duplicated purpose/summary/keyFacts wording.

  • →Reduce repeated wording across purpose, abstract, keyFacts, and sections.
Field A Field B Overlap
abstract keyFacts 30%
purpose abstract 25%

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 · 0 chars
Schema · 0 chars
AIPM · 0 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 ~34% fewer tokens than HTML-always on this pack (8 answered from machine card, 4 escalated to HTML). Descriptive only.

Orientation pack

9 questions · all layers scored

  • HTML 6/9
  • Schema 4/9
  • AIPM 8/9

Depth pack

3 questions · all layers scored

  • HTML 3/3
  • Schema 0/3
  • AIPM 0/3

Full-context pack metrics (secondary)

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

HTML

9/12 matched

4,635 full-pack tokens

Schema

4/12 matched

5,120 full-pack tokens

AIPM

8/12 matched

3,536 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) 9/12 4/12 8/12
Context size 0 chars 0 chars 0 chars
Total tokens 4,635 5,120 3,536
Tokens / correct answer 515 1,280 442
Est. cost / correct answer $0.000092 $0.000216 $0.000078
Matched per 1k tokens 1.942 0.781 2.262
Median latency 1,568 ms 1,596 ms 1,068 ms
Est. cost (USD) $0.00082 $0.00086 $0.00062

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 85.9
  • Schema 34.5
  • AIPM 100

Understanding

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

  • HTML 100
  • Schema 60
  • AIPM 80

Retrieval

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

  • HTML 100
  • Schema 0
  • AIPM 0

Evidence

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

  • HTML 79
  • Schema 36.7
  • AIPM 49.6

Metadata

Language, page kind, and freshness from page signals.

  • HTML 33.3
  • Schema 33.3
  • AIPM 100

Compression

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

  • HTML 85.9
  • Schema 34.5
  • AIPM 100

Coverage map

AIPM matched 8 question(s) (alone on Q7, Q8, Q9); HTML matched 9. HTML/Schema (or a gap) still needed on Q2, Q10, Q11, Q12. Read this as complementary coverage — AIPM orients agents cheaply; HTML supplies depth when the sidecar cannot. HTML 9/12 matched (≈515 tok/match). Schema 4/12 matched (≈1,280 tok/match). AIPM 8/12 matched (≈442 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, Q3, Q4, Q5, Q6, Q7, Q8, Q9

Alone: Q7, Q8, Q9

AIPM insufficient

Q2, Q10, Q11, Q12

HTML/Schema needed or all layers missed

Unanswered by all

—

HTML layer

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

Context fed: 0 chars

Tokens: 4,635 · median 1,568 ms

Matched this pack: 9/12

Stronger on

Understanding (100) · Retrieval (100) · Evidence (79) · Compression (85.9) · Answer Efficiency (85.9)

Weaker on

Metadata (33.3)

Schema layer

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

Context fed: 0 chars

Tokens: 5,120 · median 1,596 ms

Matched this pack: 4/12

Stronger on

—

Weaker on

Retrieval (0) · Metadata (33.3) · Evidence (36.7) · Compression (34.5) · Answer Efficiency (34.5)

AIPM layer

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

Context fed: 0 chars

Tokens: 3,536 · median 1,068 ms

Matched this pack: 8/12

Stronger on

Understanding (80) · Metadata (100) · Compression (100) · Answer Efficiency (100)

Weaker on

Retrieval (0)

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: 3,536 tok · $0.00062. Routing sim saved ~34% tokens vs HTML-always.
Deep content / research (prose facts, process detail) HTML (with optional machine orientation) HTML pack: 4,635 tok · $0.00082. Use when depth needs body prose.
Structured entity pulls (org, location, typed fields) Schema.org Schema pack: 5,120 tok · $0.00086. Dense JSON-LD tends to score well here.

Findings

  • —Of 12 questions: lowest measured retrieval cost among sufficient layers — HTML 4 · Schema 1 · AIPM 7 · multi-layer 0 · unanswered 0. Descriptive counts only; no overall ranking.
  • —Semantic Redundancy 28% — moderate field overlap; trim duplicated purpose/summary/keyFacts wording.
  • —Manifest design: Reduce repeated wording across purpose, abstract, keyFacts, and sections.
  • —Structural size is secondary. A larger AIPM is not a failure if per-question retrieval stays tiny — check Semantic Redundancy instead.
  • —AIPM matched 8 question(s) (alone on Q7, Q8, Q9); HTML matched 9. HTML/Schema (or a gap) still needed on Q2, Q10, Q11, Q12. Read this as complementary coverage — AIPM orients agents cheaply; HTML supplies depth when the sidecar cannot. HTML 9/12 matched (≈515 tok/match). Schema 4/12 matched (≈1,280 tok/match). AIPM 8/12 matched (≈442 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 the page title?

Gold: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar

Gold source: html.title · html_independent

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

Measured retrieval cost (API tokens) · lowest cost AIPM · 113 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML title 116 api 1 high yes
Schema name+headline+description+publisher+@type+block 444 api 2 low no
AIPM title 113 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: What is the page title?
  2. Planner: Planner mapped question `title` on AIPM to minimal fields: title.
  3. Slice: title — AIPM.title: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online
  4. Actual input tokens: 113
  5. LLM answer: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online
  6. Evidence: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online
  7. Confidence: high · matched yes
Execution chain · HTML · 1 round(s)
  1. Question: What is the page title?
  2. Planner: Planner mapped question `title` on HTML to minimal fields: title.
  3. Slice: title — HTML.title: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar
  4. Actual input tokens: 116
  5. LLM answer: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar
  6. Evidence: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: What is the page title?
  2. Planner: Planner mapped question `title` on SCHEMA to minimal fields: name, headline.
  3. Slice: name+headline+description+publisher+@type+block — SCHEMA.name: AigeoRadar SCHEMA.headline: Free Full LLMS.txt Generator | Create AI-Ready llms.txt On...
  4. Actual input tokens: 444
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 122 in-tok · score 0 · miss

Round 2: 322 in-tok · score 0 · miss

HTML

sufficient

Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar

score 100 · 0 chars context · 116 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 0 chars context · 444 in-tokens

Schema did not answer from 0 chars of context — additional or different layer context needed.

AIPM

sufficient

Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online

score 95 · 0 chars context · 113 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q2 · Understanding · orientation

What is the primary topic of this page?

Gold: Free Full LLMS.txt Generator – Create a Complete AI-Ready llms.txt

Gold source: html.h1 · html_independent

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

Measured retrieval cost (API tokens) · lowest cost AIPM · 408 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML h1+title+meta_description+lead_paragraphs+prose_span+body_excerpt 577 api 2 medium yes
Schema name+description+headline+publisher+@type+block 472 api 2 low no
AIPM primaryTopic+purpose+abstract+title+publisher+keyFacts+audience+contentIntent 408 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: What is the primary topic of this page?
  2. Planner: Planner mapped question `primary_topic` on AIPM to minimal fields: primaryTopic.
  3. Slice: primaryTopic+purpose+abstract+title+publisher+keyFacts+audience+contentIntent — AIPM.primaryTopic: AI visibility AIPM.purpose: This page explains what llms.txt is, why AI crawlers...
  4. Actual input tokens: 408
  5. LLM answer: AI visibility
  6. Evidence: AI visibility
  7. Confidence: low · matched no

Round 1: 104 in-tok · score 18 · miss

Round 2: 304 in-tok · score 18 · miss

Execution chain · HTML · 2 round(s)
  1. Question: What is the primary topic of this page?
  2. Planner: Planner mapped question `primary_topic` on HTML to minimal fields: h1, title.
  3. Slice: h1+title+meta_description+lead_paragraphs+prose_span+body_excerpt — HTML.h1: Free Full LLMS.txt Generator – Create a Complete AI-Ready llms.txt HTML.title: Free Full L...
  4. Actual input tokens: 577
  5. LLM answer: The primary topic of this page is a free tool to generate a complete AI-ready llms.txt file for websites.
  6. Evidence: HTML.h1: Free Full LLMS.txt Generator – Create a Complete AI-Ready llms.txt HTML.title: Free Full LLMS.txt Generator | Create AI-Ready llms...
  7. Confidence: medium · matched yes

Round 1: 140 in-tok · score 53 · match

Round 2: 437 in-tok · score 61 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: What is the primary topic of this page?
  2. Planner: Planner mapped question `primary_topic` on SCHEMA to minimal fields: name, description.
  3. Slice: name+description+headline+publisher+@type+block — SCHEMA.name: AigeoRadar SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot...
  4. Actual input tokens: 472
  5. LLM answer: The primary topic of this page is generating a complete llms.txt file for AI crawlers.
  6. Evidence: SCHEMA.name: AigeoRadar SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended...
  7. Confidence: low · matched no

Round 1: 147 in-tok · score 44 · miss

Round 2: 325 in-tok · score 44 · miss

Root cause · Layer miss

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

HTML

sufficient

The primary topic of this page is a free tool to generate a complete AI-ready llms.txt file for websites.

score 61 · 0 chars context · 577 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

partial

The primary topic of this page is generating a complete llms.txt file for AI crawlers.

score 44 · 0 chars context · 472 in-tokens

Schema was only partially aligned with gold from 0 chars of context.

AIPM

insufficient

AI visibility

score 18 · 0 chars context · 408 in-tokens

AIPM did not answer from 0 chars of context — additional or different layer context needed.

Q3 · Understanding · orientation

What is the purpose of this page?

Gold: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawlers understand your website with AigeoRadar.

Gold source: html.page_summary · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost SCHEMA · 137 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs 192 api 1 medium yes
Schema description 137 api 1 medium yes
AIPM purpose+abstract+primaryTopic+title+publisher+keyFacts+audience+contentIntent 458 api 2 medium yes
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: This page explains what llms.txt is, why AI crawlers use it, and how to generate a com...
  4. Actual input tokens: 458
  5. LLM answer: The purpose of this page is to explain what llms.txt is, why AI crawlers use it, and how to generate a complete AI-ready llms.txt file for free.
  6. Evidence: AIPM.purpose: This page explains what llms.txt is, why AI crawlers use it, and how to generate a complete AI-ready llms.txt file for free. I...
  7. Confidence: medium · matched yes

Round 1: 155 in-tok · score 50 · match

Round 2: 303 in-tok · score 50 · match

Execution chain · HTML · 1 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 — HTML.meta_description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexit...
  4. Actual input tokens: 192
  5. LLM answer: The purpose of this page is to help generate a complete llms.txt file for free to assist AI crawlers in understanding your website.
  6. Evidence: HTML.meta_description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawler...
  7. Confidence: medium · matched yes
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.
  3. Slice: description — SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity,...
  4. Actual input tokens: 137
  5. LLM answer: The purpose of this page is to help AI crawlers understand your website by generating a complete llms.txt file for free.
  6. Evidence: SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawlers u...
  7. Confidence: medium · matched yes

HTML

sufficient

The purpose of this page is to help generate a complete llms.txt file for free to assist AI crawlers in understanding your website.

score 71 · 0 chars context · 192 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

sufficient

The purpose of this page is to help AI crawlers understand your website by generating a complete llms.txt file for free.

score 73 · 0 chars context · 137 in-tokens

Schema answered using 0 chars of layer context (minimum fed this run).

AIPM

sufficient

The purpose of this page is to explain what llms.txt is, why AI crawlers use it, and how to generate a complete AI-ready llms.txt file for free.

score 50 · 0 chars context · 458 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q4 · Understanding · orientation

Summarize the page in one sentence.

Gold: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawlers understand your website with AigeoRadar.

Gold source: html.page_summary · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost AIPM · 443 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt 652 api 2 medium yes
Schema description+name+headline+publisher+@type+block 463 api 2 medium yes
AIPM abstract 443 api 2 medium yes
Execution chain · AIPM · 2 round(s)
  1. Question: Summarize the page in one sentence.
  2. Planner: Planner mapped question `abstract` on AIPM to minimal fields: abstract.
  3. Slice: abstract — AIPM.abstract: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Googl...
  4. Actual input tokens: 443
  5. LLM answer: The page offers a service to create a complete llms.txt file for free to assist AI crawlers in understanding your website.
  6. Evidence: AIPM.abstract: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawlers unders...
  7. Confidence: medium · matched yes

Round 1: 139 in-tok · score 56 · match

Round 2: 304 in-tok · score 49 · match

Execution chain · HTML · 2 round(s)
  1. Question: Summarize the 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+title+h1+prose_span+body_excerpt — HTML.meta_description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexit...
  4. Actual input tokens: 652
  5. LLM answer: The page offers a free tool to generate a complete llms.txt file to help AI crawlers like GPTBot and ClaudeBot understand and prioritize a website's content.
  6. Evidence: HTML.meta_description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawler...
  7. Confidence: medium · matched yes

Round 1: 193 in-tok · score 59 · match

Round 2: 459 in-tok · score 72 · match

Execution chain · SCHEMA · 2 round(s)
  1. Question: Summarize the page in one sentence.
  2. Planner: Planner mapped question `abstract` on SCHEMA to minimal fields: description.
  3. Slice: description+name+headline+publisher+@type+block — SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity,...
  4. Actual input tokens: 463
  5. LLM answer: AigeoRadar is a free platform that generates a complete llms.txt file and helps AI crawlers understand your website by assessing its AI readiness and authority.
  6. Evidence: SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawlers u...
  7. Confidence: medium · matched yes

Round 1: 138 in-tok · score 56 · match

Round 2: 325 in-tok · score 61 · match

HTML

sufficient

The page offers a free tool to generate a complete llms.txt file to help AI crawlers like GPTBot and ClaudeBot understand and prioritize a website's content.

score 72 · 0 chars context · 652 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

sufficient

AigeoRadar is a free platform that generates a complete llms.txt file and helps AI crawlers understand your website by assessing its AI readiness and authority.

score 61 · 0 chars context · 463 in-tokens

Schema answered using 0 chars of layer context (minimum fed this run).

AIPM

sufficient

The page offers a service to create a complete llms.txt file for free to assist AI crawlers in understanding your website.

score 56 · 0 chars context · 443 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q5 · Understanding · orientation

Who is the publisher or brand?

Gold: AigeoRadar

Gold source: html.jsonld|title · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost AIPM · 103 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML title+h1+meta_description+lead_paragraphs+prose_span+body_excerpt 540 api 2 high yes
Schema publisher+name 111 api 1 high yes
AIPM publisher 103 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: AigeoRadar
  4. Actual input tokens: 103
  5. LLM answer: AigeoRadar.
  6. Evidence: AIPM.publisher: AigeoRadar
  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: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar HTML.h1: Fr...
  4. Actual input tokens: 540
  5. LLM answer: AigeoRadar
  6. Evidence: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online — AigeoRadar
  7. Confidence: high · matched yes

Round 1: 117 in-tok · score 0 · miss

Round 2: 423 in-tok · score 100 · match

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: AigeoRadar SCHEMA.name: AigeoRadar
  4. Actual input tokens: 111
  5. LLM answer: AigeoRadar
  6. Evidence: AigeoRadar
  7. Confidence: high · matched yes

HTML

sufficient

AigeoRadar

score 100 · 0 chars context · 540 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

sufficient

AigeoRadar

score 100 · 0 chars context · 111 in-tokens

Schema answered using 0 chars of layer context (minimum fed this run).

AIPM

sufficient

AigeoRadar.

score 100 · 0 chars context · 103 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q6 · Metadata · orientation

What language is this page in?

Gold: en

Gold source: html.lang|script · html_independent

Sufficient: HTML, Schema, AIPM

Measured retrieval cost (API tokens) · lowest cost HTML · 99 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lang 99 api 1 high yes
Schema inLanguage 100 api 1 high yes
AIPM inLanguage 101 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: What language is this page in?
  2. Planner: Planner mapped question `language` on AIPM to minimal fields: inLanguage.
  3. Slice: inLanguage — AIPM.inLanguage: en
  4. Actual input tokens: 101
  5. LLM answer: The page is in English.
  6. Evidence: AIPM.inLanguage: en
  7. Confidence: high · matched yes
Execution chain · HTML · 1 round(s)
  1. Question: What language is this page in?
  2. Planner: Planner mapped question `language` on HTML to minimal fields: lang.
  3. Slice: lang — HTML.lang: lang=en
  4. Actual input tokens: 99
  5. LLM answer: The page is in English.
  6. Evidence: HTML.lang: lang=en
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 1 round(s)
  1. Question: What language is this page in?
  2. Planner: Planner mapped question `language` on SCHEMA to minimal fields: inLanguage.
  3. Slice: inLanguage — SCHEMA.inLanguage: en
  4. Actual input tokens: 100
  5. LLM answer: The page is in English.
  6. Evidence: SCHEMA.inLanguage: en
  7. Confidence: high · matched yes

HTML

sufficient

The page is in English.

score 88 · 0 chars context · 99 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

sufficient

The page is in English.

score 88 · 0 chars context · 100 in-tokens

Schema answered using 0 chars of layer context (minimum fed this run).

AIPM

sufficient

The page is in English.

score 88 · 0 chars context · 101 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q7 · Metadata · orientation

What is the content intent?

Gold: informational

Gold source: html.heuristic · html_independent

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

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

Layer Planner fields Input tokens Rounds Confidence OK?
HTML meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt 610 api 2 low no
Schema name+description+headline+publisher+@type+block 427 api 2 low no
AIPM contentIntent 101 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: What is the content intent?
  2. Planner: Planner mapped question `content_intent` on AIPM to minimal fields: contentIntent.
  3. Slice: contentIntent — AIPM.contentIntent: informational
  4. Actual input tokens: 101
  5. LLM answer: The content intent is informational.
  6. Evidence: AIPM.contentIntent: informational
  7. Confidence: high · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: What is the content intent?
  2. Planner: Planner mapped question `content_intent` on HTML to minimal fields: meta_description, lead_paragraphs.
  3. Slice: meta_description+lead_paragraphs+title+h1+prose_span+body_excerpt — HTML.meta_description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexit...
  4. Actual input tokens: 610
  5. LLM answer: The content intent is to provide a free tool for generating a complete llms.txt file to help AI crawlers understand a website.
  6. Evidence: HTML.meta_description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended, and other AI crawler...
  7. Confidence: low · matched no

Round 1: 190 in-tok · score 0 · miss

Round 2: 420 in-tok · score 0 · miss

Execution chain · SCHEMA · 2 round(s)
  1. Question: What is the content intent?
  2. Planner: Planner mapped question `content_intent` on SCHEMA to minimal fields: name, description.
  3. Slice: name+description+headline+publisher+@type+block — SCHEMA.name: AigeoRadar SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot...
  4. Actual input tokens: 427
  5. LLM answer: The content intent is to provide a free tool for generating a complete llms.txt file to help AI crawlers understand websites.
  6. Evidence: SCHEMA.name: AigeoRadar SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity, Google-Extended...
  7. Confidence: low · matched no

Round 1: 144 in-tok · score 0 · miss

Round 2: 283 in-tok · score 0 · miss

HTML

insufficient

The content intent is to provide a free tool for generating a complete llms.txt file to help AI crawlers understand a website.

score 0 · 0 chars context · 610 in-tokens

HTML did not answer from 0 chars of context — additional or different layer context needed.

Schema

insufficient

The content intent is to provide a free tool for generating a complete llms.txt file to help AI crawlers understand websites.

score 0 · 0 chars context · 427 in-tokens

Schema did not answer from 0 chars of context — additional or different layer context needed.

AIPM

sufficient

The content intent is informational.

score 88 · 0 chars context · 101 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q8 · Metadata · orientation

What page type is declared in the machine layer?

Gold: blog

Gold source: pageType · aipm_consistency

Sufficient: AIPM

Measured retrieval cost (API tokens) · lowest cost AIPM · 104 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+title+h1+meta_description+prose_span+body_excerpt 574 api 2 low no
Schema @type+name+headline+description+publisher+block 429 api 2 low no
AIPM pageType 104 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: What page type is declared in the machine layer?
  2. Planner: Planner mapped question `page_type` on AIPM to minimal fields: pageType.
  3. Slice: pageType — AIPM.pageType: blog
  4. Actual input tokens: 104
  5. LLM answer: blog
  6. Evidence: blog
  7. Confidence: high · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: What page type is declared in the machine layer?
  2. Planner: Planner mapped question `page_type` on HTML to minimal fields: lead_paragraphs.
  3. Slice: lead_paragraphs+title+h1+meta_description+prose_span+body_excerpt — HTML.lead_paragraphs: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity...
  4. Actual input tokens: 574
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 151 in-tok · score 0 · miss

Round 2: 423 in-tok · score 0 · miss

Execution chain · SCHEMA · 2 round(s)
  1. Question: What page type is declared in the machine layer?
  2. Planner: Planner mapped question `page_type` on SCHEMA to minimal fields: @type.
  3. Slice: @type+name+headline+description+publisher+block — SCHEMA.@type: Organization SCHEMA.name: AigeoRadar SCHEMA.headline: Free Full LLMS.txt Generator |...
  4. Actual input tokens: 429
  5. LLM answer: The page type declared in the machine layer is "Organization."
  6. Evidence: SCHEMA.@type: Organization SCHEMA.name: AigeoRadar SCHEMA.headline: Free Full LLMS.txt Generator | Create AI-Ready llms.txt Online SCHEMA...
  7. Confidence: low · matched no

Round 1: 103 in-tok · score 0 · miss

Round 2: 326 in-tok · score 0 · miss

HTML

skipped

UNKNOWN.

score 0 · 0 chars context · 0 in-tokens

Skipped (AIPM-only).

Schema

skipped

The page type declared in the machine layer is "Organization."

score 0 · 0 chars context · 0 in-tokens

Skipped (AIPM-only).

AIPM

sufficient

blog

score 100 · 0 chars context · 104 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q9 · Understanding · orientation

Who is the intended audience declared in AIPM?

Gold: B2B, agency, SMB

Gold source: audience · aipm_consistency

Sufficient: AIPM

Measured retrieval cost (API tokens) · lowest cost AIPM · 113 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML lead_paragraphs+title+h1+meta_description+prose_span+body_excerpt 582 api 2 low no
Schema name+description+headline+publisher+@type+block 437 api 2 low no
AIPM audience 113 api 1 high yes
Execution chain · AIPM · 1 round(s)
  1. Question: Who is the intended audience declared in AIPM?
  2. Planner: Planner mapped question `audience` on AIPM to minimal fields: audience.
  3. Slice: audience — AIPM.audience: ["B2B","agency","SMB"]
  4. Actual input tokens: 113
  5. LLM answer: The intended audience declared in AIPM includes B2B, agency, and SMB.
  6. Evidence: AIPM.audience: ["B2B","agency","SMB"]
  7. Confidence: high · matched yes
Execution chain · HTML · 2 round(s)
  1. Question: Who is the intended audience declared in AIPM?
  2. Planner: Planner mapped question `audience` on HTML to minimal fields: lead_paragraphs.
  3. Slice: lead_paragraphs+title+h1+meta_description+prose_span+body_excerpt — HTML.lead_paragraphs: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity...
  4. Actual input tokens: 582
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 152 in-tok · score 0 · miss

Round 2: 430 in-tok · score 0 · miss

Execution chain · SCHEMA · 2 round(s)
  1. Question: Who is the intended audience declared in AIPM?
  2. Planner: Planner mapped question `audience` on SCHEMA to minimal fields: name, description.
  3. Slice: name+description+headline+publisher+@type+block — SCHEMA.name: AigeoRadar SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot...
  4. Actual input tokens: 437
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 149 in-tok · score 0 · miss

Round 2: 288 in-tok · score 0 · miss

HTML

skipped

UNKNOWN.

score 0 · 0 chars context · 0 in-tokens

Skipped (AIPM-only).

Schema

skipped

UNKNOWN.

score 0 · 0 chars context · 0 in-tokens

Skipped (AIPM-only).

AIPM

sufficient

The intended audience declared in AIPM includes B2B, agency, and SMB.

score 86 · 0 chars context · 113 in-tokens

AIPM answered using 0 chars of layer context (minimum fed this run).

Q10 · Retrieval · depth

According to the page, what does it say about: "The dashboard provides a comprehensive overview …"?

Gold: The dashboard provides a comprehensive overview and updates on how well your site is optimized for AI visibility.

Gold source: html.prose · html_independent

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

Measured retrieval cost (API tokens) · lowest cost HTML · 130 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML prose_span 130 api 1 high yes
Schema description+block+name+headline+publisher+@type 626 api 2 low no
AIPM keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent 460 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: According to the page, what does it say about: "The dashboard provides a comprehensive overview …"?
  2. Planner: Planner mapped question `page_fact_1` on AIPM to minimal fields: keyFacts, sections.
  3. Slice: keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: llms.txt is designed for AI crawlers and language models. The generator creates a com...
  4. Actual input tokens: 460
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 145 in-tok · score 0 · miss

Round 2: 315 in-tok · score 0 · miss

Execution chain · HTML · 1 round(s)
  1. Question: According to the page, what does it say about: "The dashboard provides a comprehensive overview …"?
  2. Planner: Planner mapped question `page_fact_1` on HTML to minimal fields: prose_span.
  3. Slice: prose_span — HTML.prose_span: The dashboard provides a comprehensive overview and updates on how well your site i...
  4. Actual input tokens: 130
  5. LLM answer: The dashboard provides a comprehensive overview and updates on how well your site is optimized for AI visibility.
  6. Evidence: The dashboard provides a comprehensive overview and updates on how well your site is optimized for AI visibility.
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: According to the page, what does it say about: "The dashboard provides a comprehensive overview …"?
  2. Planner: Planner mapped question `page_fact_1` on SCHEMA to minimal fields: description, block.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity,...
  4. Actual input tokens: 626
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 290 in-tok · score 0 · miss

Round 2: 336 in-tok · score 0 · miss

Root cause · Depth gap (expected)

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

HTML

sufficient

The dashboard provides a comprehensive overview and updates on how well your site is optimized for AI visibility.

score 100 · 0 chars context · 130 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 0 chars context · 626 in-tokens

Schema did not answer from 0 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 0 chars context · 460 in-tokens

AIPM did not answer from 0 chars of context — additional or different layer context needed.

Q11 · Retrieval · depth

According to the page, what does it say about: "Additionally, the Action Center will list any is…"?

Gold: Additionally, the Action Center will list any issues detected along with suggested fixes, many of which can improve as your discovery assets go live.

Gold source: html.prose · html_independent

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

Measured retrieval cost (API tokens) · lowest cost HTML · 141 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML prose_span 141 api 1 high yes
Schema description+block+name+headline+publisher+@type 632 api 2 low no
AIPM keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent 466 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: According to the page, what does it say about: "Additionally, the Action Center will list any is…"?
  2. Planner: Planner mapped question `page_fact_2` on AIPM to minimal fields: keyFacts, sections.
  3. Slice: keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: llms.txt is designed for AI crawlers and language models. The generator creates a com...
  4. Actual input tokens: 466
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 148 in-tok · score 0 · miss

Round 2: 318 in-tok · score 0 · miss

Execution chain · HTML · 1 round(s)
  1. Question: According to the page, what does it say about: "Additionally, the Action Center will list any is…"?
  2. Planner: Planner mapped question `page_fact_2` on HTML to minimal fields: prose_span.
  3. Slice: prose_span — HTML.prose_span: Additionally, the Action Center will list any issues detected along with suggested...
  4. Actual input tokens: 141
  5. LLM answer: The Action Center will list any issues detected along with suggested fixes, many of which can improve as your discovery assets go live.
  6. Evidence: Additionally, the Action Center will list any issues detected along with suggested fixes, many of which can improve as your discovery assets...
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: According to the page, what does it say about: "Additionally, the Action Center will list any is…"?
  2. Planner: Planner mapped question `page_fact_2` on SCHEMA to minimal fields: description, block.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity,...
  4. Actual input tokens: 632
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 293 in-tok · score 0 · miss

Round 2: 339 in-tok · score 0 · miss

Root cause · Depth gap (expected)

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

HTML

sufficient

The Action Center will list any issues detected along with suggested fixes, many of which can improve as your discovery assets go live.

score 97 · 0 chars context · 141 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 0 chars context · 632 in-tokens

Schema did not answer from 0 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 0 chars context · 466 in-tokens

AIPM did not answer from 0 chars of context — additional or different layer context needed.

Q12 · Retrieval · depth

According to the page, what does it say about: "Another mistake is neglecting to implement the r…"?

Gold: Another mistake is neglecting to implement the recommended fixes from the Action Center, which can hinder your site's AI performance.

Gold source: html.prose · html_independent

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

Measured retrieval cost (API tokens) · lowest cost HTML · 137 input tokens

Layer Planner fields Input tokens Rounds Confidence OK?
HTML prose_span 137 api 1 high yes
Schema description+block+name+headline+publisher+@type 632 api 2 low no
AIPM keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent 466 api 2 low no
Execution chain · AIPM · 2 round(s)
  1. Question: According to the page, what does it say about: "Another mistake is neglecting to implement the r…"?
  2. Planner: Planner mapped question `page_fact_3` on AIPM to minimal fields: keyFacts, sections.
  3. Slice: keyFacts+purpose+abstract+primaryTopic+title+publisher+audience+contentIntent — AIPM.keyFacts: llms.txt is designed for AI crawlers and language models. The generator creates a com...
  4. Actual input tokens: 466
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 148 in-tok · score 0 · miss

Round 2: 318 in-tok · score 0 · miss

Execution chain · HTML · 1 round(s)
  1. Question: According to the page, what does it say about: "Another mistake is neglecting to implement the r…"?
  2. Planner: Planner mapped question `page_fact_3` on HTML to minimal fields: prose_span.
  3. Slice: prose_span — HTML.prose_span: Another mistake is neglecting to implement the recommended fixes from the Action Ce...
  4. Actual input tokens: 137
  5. LLM answer: Another mistake is neglecting to implement the recommended fixes from the Action Center, which can hinder your site's AI performance.
  6. Evidence: Another mistake is neglecting to implement the recommended fixes from the Action Center, which can hinder your site's AI performance.
  7. Confidence: high · matched yes
Execution chain · SCHEMA · 2 round(s)
  1. Question: According to the page, what does it say about: "Another mistake is neglecting to implement the r…"?
  2. Planner: Planner mapped question `page_fact_3` on SCHEMA to minimal fields: description, block.
  3. Slice: description+block+name+headline+publisher+@type — SCHEMA.description: Generate a complete llms.txt file for free. Help GPTBot, ClaudeBot, Perplexity,...
  4. Actual input tokens: 632
  5. LLM answer: UNKNOWN.
  6. Evidence: —
  7. Confidence: low · matched no

Round 1: 293 in-tok · score 0 · miss

Round 2: 339 in-tok · score 0 · miss

Root cause · Depth gap (expected)

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

HTML

sufficient

Another mistake is neglecting to implement the recommended fixes from the Action Center, which can hinder your site's AI performance.

score 100 · 0 chars context · 137 in-tokens

HTML answered using 0 chars of layer context (minimum fed this run).

Schema

insufficient

UNKNOWN.

score 0 · 0 chars context · 632 in-tokens

Schema did not answer from 0 chars of context — additional or different layer context needed.

AIPM

insufficient

UNKNOWN.

score 0 · 0 chars context · 466 in-tokens

AIPM did not answer from 0 chars of context — additional or different layer context needed.

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 12:54 PM · aipm_benchmark_score_v10

Share

Per-question context chain across layers.

Post on X LinkedIn WhatsApp
About the lab →