AA-Omniscience (Accuracy)
Accuracy component of AA-Omniscience, Artificial Analysis's benchmark of factual knowledge and hallucination behaviour.
Actively evaluated by Artificial Analysis on current frontier models; no saturation evidence recorded.
Performance Timeline
Longitudinal progression of model scores against human baselines.Performance & Historical Trajectory
Empirical score progression across model release dates and evaluation rounds.
| Model | Score | Date | Source Type | Provenance |
|---|---|---|---|---|
| NVIDIA Nemotron 3 Nano 4B | 9.5% | 2026-11-15 | independent | Source ↗ |
| DeepSeek V4.1 Flash 1017 | 39.6% | 2026-11-15 | independent | Source ↗ |
| Muse Glimmer | 28% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3 Ultra | 21.2% | 2026-11-15 | independent | Source ↗ |
| Solar Mini 4 | 18.2% | 2026-11-15 | independent | Source ↗ |
| Kimi K3.1 Pro | 48.9% | 2026-11-15 | independent | Source ↗ |
| Beam 1.0 | 30.7% | 2026-11-15 | independent | Source ↗ |
| Inkling Small | 31.8% | 2026-11-15 | independent | Source ↗ |
| Gemini 3.8 Flash | 50.5% | 2026-11-15 | independent | Source ↗ |
| Muse Ember | 21.1% | 2026-11-15 | independent | Source ↗ |
| Ling 3.0 Tiny | 37.9% | 2026-11-15 | independent | Source ↗ |
| A.X-K2 | 17.5% | 2026-11-15 | independent | Source ↗ |
| DeepSeek V4-Pro | 45.2% | 2026-11-15 | independent | Source ↗ |
| Gemini 4 Hydrogen | 50.5% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8 2.4T A95B | 31.4% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8-Flash-Next | 23.5% | 2026-11-15 | independent | Source ↗ |
| Ling 3.1 Flash 1019 | 26.5% | 2026-11-15 | independent | Source ↗ |
| Granite 4.2 8B | 11.8% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8 Max (0902) | 29.5% | 2026-11-15 | independent | Source ↗ |
| JT-4.1 Flash 236B A21B | 37.9% | 2026-11-15 | independent | Source ↗ |
| K2 Horizon 375B A23B | 18.6% | 2026-11-15 | independent | Source ↗ |
| gpt-oss-120b | 21.7% | 2026-11-15 | independent | Source ↗ |
| GLM 5.3 | 32.3% | 2026-11-15 | independent | Source ↗ |
| Gemma 4 31B | 19.9% | 2026-11-15 | independent | Source ↗ |
| Hy4 | 50% | 2026-11-15 | independent | Source ↗ |
| MiniMax M3.1 Flash | 45.1% | 2026-11-15 | independent | Source ↗ |
| Agnes 2.5 Pro Theta | 21.3% | 2026-11-15 | independent | Source ↗ |
| Mistral Medium 3.5 | 23% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3.5 Super VL | 21.2% | 2026-11-15 | independent | Source ↗ |
| Inkling 2 Small | 33.8% | 2026-11-15 | independent | Source ↗ |
| Muse Wonder Full | 41% | 2026-11-15 | independent | Source ↗ |
| Ring-2.6-1T | 26.3% | 2026-11-15 | independent | Source ↗ |
| GPT-6 Luna | 39.6% | 2026-11-15 | independent | Source ↗ |
| Apodex 1.1 | 28.2% | 2026-11-15 | independent | Source ↗ |
| GPT-6.1 Sol | 61.8% | 2026-11-15 | independent | Source ↗ |
| Gemini 3.5 Flash-Lite | 26.5% | 2026-11-15 | independent | Source ↗ |
| Step 5 | 36.4% | 2026-11-15 | independent | Source ↗ |
| Granite 4.2 3B | 9.2% | 2026-11-15 | independent | Source ↗ |
| Claude Fable 5.2 | 79.1% | 2026-11-15 | independent | Source ↗ |
| GPT-6.1 Terra | 47.6% | 2026-11-15 | independent | Source ↗ |
| Gemini 4 Carbon | 58.3% | 2026-11-15 | independent | Source ↗ |
| Inkling | 39.3% | 2026-11-15 | independent | Source ↗ |
| Granite 4.2 30B | 11.2% | 2026-11-15 | independent | Source ↗ |
| MiMo-V2.6-Flash | 27.2% | 2026-11-15 | independent | Source ↗ |
| Mistral Small 4 | 21.4% | 2026-11-15 | independent | Source ↗ |
| Solar Pro 4 | 18.4% | 2026-11-15 | independent | Source ↗ |
| K-EXAONE 2.0 | 14.5% | 2026-11-15 | independent | Source ↗ |
| GLM 5.3 Flash | 28.2% | 2026-11-15 | independent | Source ↗ |
| Claude Sonnet 5.5 | 51.5% | 2026-11-15 | independent | Source ↗ |
| North Mini Code | 18.4% | 2026-11-15 | independent | Source ↗ |
| GPT-6 Astra | 61.6% | 2026-11-15 | independent | Source ↗ |
| Claude Haiku 5.5 | 34.7% | 2026-11-15 | independent | Source ↗ |
| Grok 4.7 | 42.7% | 2026-11-15 | independent | Source ↗ |
| MiMo-V2.6-Pro | 34.8% | 2026-11-15 | independent | Source ↗ |
| MiniMax M3 | 44.9% | 2026-11-15 | independent | Source ↗ |
| Mistral Large 4 | 26% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3.5 Lightning | 14.7% | 2026-11-15 | independent | Source ↗ |
| Command A+ | 9.8% | 2026-11-15 | independent | Source ↗ |
| Claude Opus 5.5 | 65.5% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8 27B | 17.3% | 2026-11-15 | independent | Source ↗ |
| Muse Spark Open | 42% | 2026-11-15 | independent | Source ↗ |
| Muse Wonder | 38% | 2026-11-15 | independent | Source ↗ |
| Gemini 4 Argon | 48.3% | 2026-11-15 | independent | Source ↗ |
| Nanbeige4.3-33B | 11.7% | 2026-11-15 | independent | Source ↗ |
Human Baseline & Difficulty Horizon
Calibrated human reference points, specialist benchmarks, and ceiling thresholds.No peer-reviewed human baseline has been published for this benchmark.
Metric & Scoring Methodology
Verification protocols, aggregation formulas, and specialized metric variants.accuracy (%)Dataset & Compute Cost
Evaluation volume, public availability, API pricing, and local hardware requirements.$5 – $20 USD for full benchmark evaluation run on frontier APIs.
1x NVIDIA RTX 4090 (24GB) or A100 (40GB/80GB) via vLLM / SGLang
How to Run & Reproduce
Standardized evaluation protocols, CLI commands, and reproducible runner templates.lm_eval --model hf --model_args pretrained=<model_path> --tasks aa-omniscience-accuracy --batch_size autoopencompass --datasets aa-omniscience-accuracy --models <model_config># Standard API Evaluation Loop
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="gpt-4o",
messages=[{"role": "user", "content": prompt}],
temperature=0.0,
)When publishing results for AA-Omniscience (Accuracy), always report the exact prompt template, few-shot exemplar ordering, sampling temperature (temperature=0), maximum reasoning budget tokens, and the precise timestamped model snapshot ID.
Contamination & Memorization Analysis
Audit of pretraining exposure risks, memorization vectors, and refresh policies.Static fixed snapshot
Public on web / HuggingFace