# Model Documentation Evaluation Template # Metadata model_name: "GPT-5.5" model_link: "https://openai.com/index/introducing-gpt-5-5/" organization: "OpenAI" org_link: "https://openai.com/" evaluation_date: "2026-07-09" public_summary_link: "cdn.openai.com/pdf/eb1f5ac3-009e-4d51-b2da-f5f327913115/gpt-5-5-eu-ai-act-public-summary-of-training-content.pdf" public_summary_date: "2026-06-26" public_summary_location: "https://help-lb.openai.com/en/articles/12141645-eu-ai-act" model_publication_date: "2025-08-07" category: "New model" # New model, fine-tuned model archive_file_name: "GPT_5_5_2026_07_14.pdf" # Requirement assessments # Each requirement should be assessed once with a score from 0-10 # Leave blank for N/A, and assign a max score of 0 S1: # Document-level requirements D1: # Clarity score: 9 max_score: 11 notes: "Document does not indicate whether it is the latest version or outdated. Document does not provide link to other versions of the document." D2: # Completeness score: 4 max_score: 4 notes: "" D3: # Consistency score: 7 max_score: 7 notes: "" D4: # Correctness score: 3.5 max_score: 5 notes: "Document does not provide a link to the authoritative source of the document. Date format slightly incorrect." D5: # Accessibility score: 6.5 max_score: 10 notes: "Document is only linked in a single location; not very discoverable. Document is not provided in the same context as the model." D6: # Comprehension score: 6 max_score: 9 notes: "Document does not clearly indicate changes from previous version. Document does not indicate where notice of updates or changes will be provided." S2: # General information D1: # Clarity score: 19 max_score: 40 notes: "Contact details refer to general business address, generic contact point. Description of types of content describes goals for which content is included. Languages are not described exactly (i.e. as a list), EU languages are not fully mentioned." D2: # Completeness score: 49 max_score: 72 notes: "No links to additional publicly available documentation are provided with versioned model name. National, regional, demographic specificities of training data not provided." D3: # Consistency score: 8 max_score: 8 notes: "" D4: # Correctness score: 17 max_score: 18 notes: "Latest data of data acquisition/collection for model training is not provided exactly." D5: # Accessibility score: 0 max_score: 0 notes: "" D6: # Comprehension score: 10 max_score: 11 notes: "It is not indicated whether the model is continuously trained on new or dynamic data after the latest data of data acquisition/collection." S3: # Public datasets D1: # Clarity score: 0 max_score: 92 notes: "List of large publicly available datasets only describes CommonCrawl. General description of other publicly available information contains some unnecessary information." D2: # Completeness score: 91 max_score: 111 notes: "List of large publicly available datasets only describes CommonCrawl. Descriptions do not contain info relating to approximate start and end dates of data collection." D3: # Consistency score: 1 max_score: 3 notes: "List of large publicly available datasets only describes CommonCrawl." D4: # Correctness score: 23 max_score: 23 notes: "" D5: # Accessibility score: 0 max_score: 20 notes: "List of large publicly available datasets only describes CommonCrawl." D6: # Comprehension score: 3 max_score: 19 notes: "Descriptions provided make use of language which is difficult to understand by someone without expertise." S4: # Private datasets D1: # Clarity score: 0 max_score: 32 notes: "General description of non-publicly known private datasets obtained from third parties contains some superfluous information." D2: # Completeness score: 24 max_score: 88 notes: "General description of non-publicly known private datasets obtained from third parties lack recommended information." D3: # Consistency score: 3 max_score: 3 notes: "" D4: # Correctness score: 24 max_score: 24 notes: "" D5: # Accessibility score: 0 max_score: 0 notes: "" D6: # Comprehension score: 16 max_score: 16 notes: "" S5: # Scraped/crawled data D1: # Clarity score: 122 max_score: 129 notes: "Exact dates of data collection are not given." D2: # Completeness score: 94 max_score: 158 notes: "Suggested information is not provided in the comprehensive description of the type of content and online sources crawled." D3: # Consistency score: 3 max_score: 3 notes: "" D4: # Correctness score: 61 max_score: 93 notes: "Dates of data collection are not given in the appropriate format." D5: # Accessibility score: 0 max_score: 25 notes: "Summary of the most relevant crawled domains not properly provided." D6: # Comprehension score: 20 max_score: 58 notes: "Comprehensive description of the type of content and online sources crawled contains some superfluous information." S6: # User data D1: # Clarity score: 8.5 max_score: 16 notes: "General description of provider's services or products that were used to collect user data contains some links to external pages." D2: # Completeness score: 19.5 max_score: 27 notes: "General description of provider's services or products that were used to collect user data are found in various locations." D3: # Consistency score: 3 max_score: 3 notes: "" D4: # Correctness score: 12 max_score: 12 notes: "" D5: # Accessibility score: 15 max_score: 15 notes: "" D6: # Comprehension score: 0 max_score: 15 notes: "General description of provider's services or products that were used to collect user data are found in various locations." S7: # Synthetic & other D1: # Clarity score: 59 max_score: 65 notes: "Listed model used to general synthetic data is not necessarily uniquely identifiable" D2: # Completeness score: 41 max_score: 47 notes: "Models used to general synthetic data is not necessarily exhaustive." D3: # Consistency score: 6 max_score: 6 notes: "" D4: # Correctness score: 47 max_score: 47 notes: "" D5: # Accessibility score: 17 max_score: 17 notes: "" D6: # Comprehension score: 12 max_score: 24 notes: "Information about synthetic models does not include a general description of the training data of the models." S8: # Data processing D1: # Clarity score: 18 max_score: 18 notes: "" D2: # Completeness score: 10.5 max_score: 18 notes: "General description of measures taken to remove illegal content describes measures as 'may includ[ing]' certain measures, not will including." D3: # Consistency score: 3 max_score: 3 notes: "" D4: # Correctness score: 36 max_score: 36 notes: "" D5: # Accessibility score: 15 max_score: 15 notes: "" D6: # Comprehension score: 33 max_score: 69 notes: "Provider does not describe how they use datasets with opt-out mechanisms, nor how they ensure that they are up to date with user rights requests." general_notes: "The public summary for GPT-5.5 can be found on OpenAI's site, linked on one of OpenAI's help pages. The summary matches the structure of the template, with points being lost primarily due to providing high-level information where more detailed information is encouraged. We assessed the summary's score to be 62.52% with Grade C+ for transparency, and 47.52% with Grade D+ for usefulness."