Overview of Quality Assessment Findings

This page provides an overview of our assessment of 64 public summaries. Each summary is scored for two dimensions: Transparency (T) on how the information is provided, and Usefulness (U) for whether it is sufficient for stakeholders' needs. For convenience, all scores are reflected as a percentage (out of hundred) and the grades are expressed on a scale from A+ (highest) to F (lowest). Sections in public summaries that were not filled in because they were not applicable are marked as "N/A". You can click on the model name for a detailed description of the scores and our notes. The evaluation steps are described on our methodology page along with a FAQ.

Note that this page does not display models whose summaries we are currently evaluating nor models for which a public summary is currently missing.

Section→ Grade Overall General Information Public Data Sources Private Data Sources Scraped/Crawled Data User Data Synthetic & Other Data Data Processing Document
Model↓ T U T U T U T U T U T U T U T U T U T U
Baguettotron and Monad A+ A+ 97 97 92 100 100 100 100 N/A 100 N/A 100 N/A 100 100 100 100 88 76
Pleias 1.0 A   A+ 94 97 100 93 100 100 100 N/A 88 100 100 N/A 100 100 100 100 88 76
Apertus A   A+ 90 97 68 93 100 100 100 N/A 100 N/A 100 N/A 100 N/A 100 100 85 84
Luciole Base A   A+ 90 95 95 93 85 100 100 N/A 100 N/A 100 N/A 82 N/A 100 100 92 68
Apertus v1.5 B+ A+ 86 96 68 93 100 100 100 N/A 80 100 100 N/A 100 N/A 100 100 87 68
FIBO B+ A+ 85 96 93 100 100 N/A 63 100 100 N/A 100 N/A 100 N/A 100 100 90 76
Bria 3.2 B+ A+ 86 95 69 92 100 N/A 97 100 100 N/A 100 N/A 100 N/A 100 100 72 76
Luciole Instruct 1.1 A   B+ 90 88 82 93 92 100 100 N/A 100 N/A 100 N/A 89 0 100 100 92 68
SmolLM3-3B B+ B+ 81 85 68 90 96 100 100 N/A 100 N/A 100 N/A 60 0 92 100 70 56
Domyn Large B+ B+ 83 81 80 90 76 100 85 N/A 83 N/A 78 N/A 97 N/A 98 77 79 57
Kolibri-1 A+ C+ 96 66 85 100 100 100 100 N/A 100 N/A 100 N/A 100 N/A 100 39 87 88
Bielik v3 11B Instruct B+ C+ 89 71 84 93 92 100 100 N/A 83 51 100 N/A 100 100 100 71 88 84
PLLuM-12B-base-2512 B+ C+ 86 68 89 100 94 100 100 N/A 75 51 100 N/A 100 100 100 57 88 84
Adobe Firefly C+ B+ 71 82 71 92 81 100 16 N/A 78 100 100 N/A 89 0 100 72 74 68
Nemotron 3 and 3.5 Family B   C+ 75 73 82 94 69 100 66 44 65 100 100 N/A 96 41 100 57 85 84
Inkling C+ C+ 67 70 72 100 54 100 89 100 56 39 100 N/A 63 58 100 78 85 81
Inkling Small C+ C+ 67 69 72 100 49 100 89 100 56 39 100 N/A 70 58 100 78 85 68
FLUX.3 B   C   75 58 66 100 100 100 61 100 100 N/A 45 75 89 50 75 21 85 68
Nova 2 Lite C+ C+ 71 62 82 100 51 44 78 100 76 39 45 50 78 58 80 78 79 84
Nemotron Nano 12B v2 C+ C+ 62 70 79 94 66 100 46 44 52 39 100 N/A 57 0 100 100 62 84
Grok Voice Think Fast C+ C   71 55 76 92 64 48 84 100 64 51 100 N/A 64 58 100 35 85 65
FastwebMIIA C   C+ 53 73 74 90 32 44 45 44 33 N/A 100 N/A 40 50 92 100 74 68
Minimax M3 C+ C   70 53 64 100 52 41 50 100 74 39 74 50 92 100 98 35 88 81
MAI-Image-2.6 C+ C   70 50 82 100 54 44 23 100 80 39 100 100 100 0 75 26 85 81
MAI Cyber 1 Flash C+ C   70 50 82 100 58 44 23 100 80 39 100 100 79 0 75 26 83 81
MAI Code 1.1 Flash C+ C   70 50 82 100 54 44 23 100 80 39 100 100 100 0 75 26 83 81
MAI Code 1 Flash C+ C   69 50 82 100 54 44 23 100 80 39 100 100 79 0 75 26 83 81
MAI-Image-2.5 C+ D+ 71 46 74 100 35 44 89 100 80 39 100 N/A 65 N/A 75 26 83 81
Hy3 C+ C   65 52 59 92 54 44 89 100 57 39 100 N/A 57 0 100 57 85 65
MAI-Image-2 C+ D+ 70 46 74 100 29 44 89 100 80 39 100 N/A 65 N/A 75 26 83 81
Ministral 3 14B C   C   58 58 74 90 48 44 66 100 40 39 72 50 57 0 100 78 88 81
Ministral 3 8B C   C   58 58 74 90 48 44 66 100 40 39 72 50 57 0 100 78 88 81
Ministral 3 3B C   C   58 58 74 90 48 44 66 100 40 39 72 50 57 0 100 78 88 81
GPT-5.5 C+ D+ 67 47 66 90 50 7 34 100 73 24 74 50 92 70 90 57 85 65
GPT-5.6 Luna C+ D+ 66 47 66 90 46 7 34 100 73 24 74 50 92 70 90 57 85 65
Gemma 4 C+ C   61 53 72 54 6 0 67 100 64 39 45 50 78 58 90 67 85 57
GPT-6 Astra C+ D+ 66 47 66 90 46 7 34 100 73 24 74 50 89 70 90 57 85 65
GPT-5.2 C+ D+ 66 47 66 90 46 7 34 100 73 24 74 50 89 70 90 57 85 65
Gemini 3 Pro C+ C   60 53 72 54 6 0 67 100 61 39 45 50 78 58 90 67 85 65
Mistral Small 4 C   C   54 58 71 90 47 44 66 100 40 39 72 50 26 0 100 78 88 81
Mistral Large 3 C   C   54 58 71 90 47 44 66 100 40 39 72 50 26 0 100 78 88 81
GPT-5.4 Nano C+ D+ 64 47 66 90 46 7 34 100 73 24 74 50 78 70 90 57 85 65
GPT Images 2.0 C+ D+ 69 42 66 90 46 7 92 N/A 73 24 74 50 85 58 90 57 85 65
Claude Haiku 5.5 C+ D+ 60 49 67 93 41 44 56 100 69 24 45 50 56 41 90 57 85 68
Grok 4.5 C   C   57 51 77 100 7 0 34 100 62 39 100 100 62 58 76 35 85 65
Claude Mythos 5.1 / Fable 5.1 C   D+ 58 49 67 93 41 44 34 100 69 24 45 50 56 41 90 57 85 68
Command A+ C   C   55 51 75 78 41 44 66 100 37 24 61 50 67 58 90 57 85 84
Claude Mythos 5 / Claude Fable 5 C   D+ 57 48 71 78 36 44 34 100 68 24 45 50 51 41 90 57 85 76
Claude Mythos Preview C   D+ 57 48 71 78 36 44 34 100 68 24 45 50 51 41 90 57 85 76
Claude Opus 5 C   D+ 57 48 71 78 36 44 34 100 68 24 45 50 51 41 90 57 85 68
Claude Opus 4.8 C   D+ 57 48 71 78 36 44 34 100 68 24 45 50 51 41 90 57 85 68
Claude Sonnet 5 C   D+ 57 48 71 78 36 44 34 100 68 24 45 50 51 41 90 57 85 68
Claude Opus 4.7 C   D+ 57 48 71 78 36 44 34 100 68 24 45 50 51 41 90 57 85 68
Muse Glimmer C   D+ 53 49 68 90 32 44 72 100 55 24 86 100 15 0 66 48 75 57
Muse Image C   D+ 53 49 67 90 32 44 72 100 55 24 72 100 15 0 66 48 88 65
Muse Spark C   D+ 53 48 67 90 32 44 72 100 55 24 86 100 15 0 66 48 77 50
Phi 4 Reasoning Vision 15B C+ D   63 39 69 100 42 44 100 N/A 100 N/A 100 N/A 100 100 15 0 87 81
Seed 2.0 Pro C   D+ 54 44 65 92 20 0 56 100 68 39 33 0 56 58 76 57 85 65
Seedance 2.5 C   D+ 52 41 65 92 20 0 56 100 64 24 100 N/A 36 0 76 57 85 65
Seedance 2.0 C   D+ 51 41 65 92 20 0 56 100 61 24 100 N/A 36 0 76 57 85 65
Seedream 5.0 Pro C   D   52 39 61 57 20 0 56 100 64 24 100 N/A 36 0 76 57 85 65
DeepSeek-V4 D+ D+ 43 41 65 57 54 44 56 100 0 0 100 N/A 56 58 100 57 87 68
Midjourney Image and Video family of models D   D+ 38 45 71 64 10 44 23 100 32 24 100 N/A 19 0 100 57 83 65
Phi-4 D   F   32 24 60 100 3 0 53 N/A 0 N/A 0 N/A 8 0 52 0 85 81