Best AI Models for Knowledge
Ranking of language models by factual accuracy and knowledge depth. These models minimize hallucinations and provide reliable information.Updated August 2026. Based on independent benchmark data.
Why This Matters
Choosing the right model for knowledge is about more than raw benchmark scores. The models below are ranked using a weighted blend of category-specific benchmarks, arena Elo, and the intelligence index — but real-world performance depends on your context window, latency budget, and integration path.
For knowledge, we weighted the benchmarks that most strongly predict success in production: task accuracy first, then speed and cost efficiency. Leaders on this page consistently outperform on both standardized tests and real-world workloads.
If you operate under a strict budget, check the Best Value LLMs ranking. If latency matters more than raw quality, see Fastest LLMs. Both pages use the same underlying dataset with different weightings.
| # | Model | Provider | Score | Speed | Price (output) |
|---|---|---|---|---|---|
| 1 | Claude Mythos Preview | Anthropic | 100.0 | 80 t/s | $15/M |
| 2 | Gemini 3.1 Pro | 97.6 | 136 t/s | $12/M | |
| 3 | Gemini 3 Pro | 92.9 | 120 t/s | $5/M | |
| 4 | Claude Mythos 5 | Anthropic | 75.0 | 50 t/s | $50/M |
| 5 | Qwen3.7 Max | Alibaba | 69.3 | 203 t/s | $7.5/M |
| 6 | Claude Sonnet 4.6 (max) | Anthropic | 67.0 | 50 t/s | $15/M |
| 7 | Qwen3.8 Flash Next | Alibaba | 64.0 | 168 t/s | $0.4/M |
| 8 | GLM-5.3 Flash | Z.ai | 60.5 | 182 t/s | $0.3/M |
| 9 | Claude Fable 5.1 | Anthropic | 58.0 | 85 t/s | $15/M |
| 10 | Gemini 3.8 Flash | 58.0 | 302 t/s | $3.75/M | |
| 11 | Gemini 3.8 Flash (Medium Reasoning) | 58.0 | 315 t/s | $3.75/M | |
| 12 | Gemini 3.8 Flash (Low Reasoning) | 58.0 | 340 t/s | $3.75/M | |
| 13 | Grok 4.7 (high) | xAI | 55.0 | 80 t/s | $15/M |
| 14 | Grok 4.7 | xAI | 55.0 | 95 t/s | $15/M |
| 15 | Granite 4.2 30B Instruct | IBM | 52.0 | 78 t/s | $1.2/M |
| 16 | GPT-5.5 Pro | OpenAI | 50.0 | 60 t/s | $10/M |
| 17 | Qwen3.6 Plus | Alibaba | 48.4 | 55 t/s | $3/M |
| 18 | Agnes 2.5 Pro Beta | Agnes | 48.0 | 65 t/s | $3/M |
| 19 | DeepSeek V4.1 Flash | DeepSeek | 46.0 | 214 t/s | $1.2/M |
| 20 | Claude Fable 5.1 (XHigh) | Anthropic | 45.0 | 78 t/s | $15/M |
Top 5 Knowledge — Quick Comparison
Methodology
Rankings are based on independent benchmark evaluations. The knowledge score is computed from multiple standardized benchmarks and normalized to 0–100. Data is updated regularly from public benchmark results.
Related Rankings
Frequently Asked Questions
Which AI model has the highest factual accuracy?
Claude Opus 5, GPT-5, and Gemini 2.0 Pro demonstrate the highest factual recall with minimal hallucination rates.