Best LLMs for Knowledge — October 2026 Leaderboard
Data refreshed:
General knowledge and factual understanding
As of October 2026, the top knowledge model on the BenchLM leaderboard is Claude Opus 5.5 with a BenchAlign knowledge score of 89.5.
Decision lens: the score determines position; Supported and Estimated labels describe the evidence behind that position without removing sparsely reported models.
- Data refreshed
- October 29, 2026
- Ranked
- 179 of 1004 models
- Supported / Estimated
- 89 / 90
- Scored evidence
- 10 of 23 benchmarks
23 tracked benchmarks
MMLU, GPQA, GPQA-D, SuperGPQA, MMLU-Pro, HLE, HLE-Diamond, HLE-Diamond (highest effort), HLE-Diamond (web+code), FrontierScience, HLE w/o tools, SimpleQA, HealthBench Hard, HealthBench Professional, MedXpertQA (Text), FrontierScience Research, MMLU-Pro (Arcee), MMMLU
Scope: Frontier science, Broad academic knowledge, Factuality
Evidence set: MMLU, GPQA, GPQA-D, SuperGPQA, MMLU-Pro, HLE, HLE-Diamond, HLE-Diamond (highest effort), HLE-Diamond (web+code), FrontierScience, HLE w/o tools, SimpleQA, HealthBench Hard, HealthBench Professional, MedXpertQA (Text), FrontierScience Research, MMLU-Pro (Arcee), MMMLU
Scope: Frontier science, Broad academic knowledge, Factuality
Best Knowledge picks
BenchLM summaries for knowledge plus the practical tradeoffs users check next: open weights, price, speed, latency, and context.
Knowledge Leaderboard
Primary score: BenchAlign knowledge score. Higher values rank first. Use the Show metric control to change the value shown in each row.
Filters
| Parameters (B) | ||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Not reported | Closed | 89.5% | 89.53 | — | — | — | — | — | — | — | — | — | — | 64.4% | — | — | 65.6% | — | — | — | — |
2 | Not reported | Closed | 88.2% | 88.24 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
3 | Not reported | Closed | 86.2% | 86.18 | — | — | — | — | — | — | — | — | — | — | — | — | 36.2% | 64.2% | — | — | — | — |
4 | Not reported | Closed | 86.0% | 86.04 | — | — | — | — | — | 65% | — | — | — | — | 60.9% | — | — | — | — | — | — | — |
5 | Not reported | Closed | 85.8% | 85.76 | — | 96% | 96.0% | — | — | — | — | — | — | — | — | — | 36.6% | 64.7% | — | — | — | — |
6 | Not reported | Closed | 82.6% | 82.61 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
7 | Not reported | Closed | 81.7% | 81.65 | — | — | — | — | — | — | — | — | — | — | 56.9% | — | — | 69.2% | — | — | — | — |
8 | Not reported | Closed | 81.5% | 81.46 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
9 | Not reported | Closed | 81.4% | 81.4 | — | — | — | — | — | 64.7% | — | — | — | — | 56.3% | — | — | 59.8% | — | — | — | — |
10 | Not reported | Closed | 80.2% | 80.16 | — | — | — | — | — | — | — | — | — | — | — | — | 30.1% | 60.8% | — | — | — | — |
11 | Not reported | Closed | 77.1% | 77.12 | — | — | — | — | — | — | — | — | — | — | — | — | 31.4% | 62.7% | — | — | — | — |
12 | Not reported | Closed | 76.3% | 76.32 | — | 94.6% | 94.6% | — | — | — | — | — | — | — | — | — | 33.1% | 60.5% | — | — | — | — |
13 | Not reported | Pending | 75.1% | 75.13 | — | 94.2% | 94.2% | — | — | 70% | — | — | — | — | 51.1% | — | — | — | — | — | — | — |
14 | Not reported | Closed | 74.2% | 74.21 | — | — | — | — | — | — | — | — | — | — | — | — | — | 62.3% | — | — | — | — |
15 | Not reported | Closed | 71.2% | 71.23 | — | 93.6% | 93.6% | — | — | 52.2% | — | — | — | — | 41.4% | — | — | — | — | — | — | — |
16 | Not reported | Closed | 70.9% | 70.93 | — | 93.6% | 93.6% | — | — | 57.9% | — | — | — | — | 49.8% | — | — | — | — | — | — | — |
17 | Not reported | Closed | 70.8% | 70.77 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
18 | Not reported | Closed | 69.2% | 69.19 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
19 | Not reported | Closed | 69.1% | 69.1 | — | 92.9% | 92.9% | — | — | — | — | — | — | — | — | — | 32.7% | 57.7% | — | — | — | — |
20 | Not reported | Closed | 68.7% | 68.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | 60.5% | — | — | — | — |
21 | Not reported | Closed | 68.5% | 68.47 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
22 | Not reported | Closed | 68.1% | 68.1 | — | 92.8% | 92.8% | — | — | 52.1% | — | — | — | — | 39.8% | — | 40.1% | 48.1% | 59.6% | — | — | — |
23 | Not reported | Closed | 67.6% | 67.6 | — | — | — | — | — | 62.1% | — | — | — | — | 52.2% | — | — | 59.3% | — | — | — | — |
| Not reported | Open | 67.2% | 67.22 | — | 93.5% | 93.5% | — | — | 56% | — | — | — | — | 43.5% | — | — | — | — | — | — | — | |
25 | Not reported | Closed | 66.4% | 66.44 | — | — | 94.3% | — | — | — | — | — | — | — | 45.4% | — | 20.6% | — | 71.5% | — | — | — |
Query current rankings and available benchmark results from your scripts or AI assistant. Start with 1,000 free reads each month.
See queries and coverageTop AI models for Knowledge — October 2026
As of October 2026, Claude Opus 5.5 leads the BenchAlign knowledge leaderboard with a score of 89.5, followed by Gemini 4 Carbon (88.2) and GPT-6.1 Sol (86.2). BenchLM is currently showing 89 Supported and 90 Estimated models in this category.
Ranks #1 on the current knowledge board with a Supported evidence label.
Ranks #2 on the current knowledge board with a Supported evidence label.
Ranks #3 on the current knowledge board with a Supported evidence label.
What changed
Claude Opus 5.5 ranks #1 at 89.5 with a Supported evidence label.
Gemini 4 Carbon ranks #2 at 88.2 with a Supported evidence label.
GPT-6.1 Sol ranks #3 at 86.2 with a Supported evidence label.
Top models by benchmark
Expert-level questions in biology, physics, and chemistry(7% of category score)
Score in Context
What these scores mean
BenchAlign places direct benchmarks and independent external signals on a common calibrated scale. The score is relative to the current evidence universe; it is not a raw percentage from any single test.
Known limitations
The table orders point estimates. Conditional score ranges describe uncertainty under the scoring assumptions; their coverage after category changes and their ability to establish rank confidence have not been validated. Supported and Estimated describe the evidence behind each row.
How we weight
Knowledge has no external category source, so admitted benchmark protocols carry it. Where category evidence is thin, the score pools toward the model's general capability, estimated without human-preference leaderboards. Evidence sources are calibrated for difficulty before aggregation, and no generated benchmark row contributes to the score.
Leaderboards exclude benchmark rows that BenchLM generated from other scores or cloned from reference models. When a weighted benchmark is missing after that filter, the category falls back to the remaining trustworthy public rows instead of filling the gap with synthetic values.
The full scoring rules, freshness handling, and runtime/pricing caveats live on the BenchLM methodology page.
Scroll horizontally to read the full evidence ledger.
| Benchmark | Weight | Status | Description |
|---|---|---|---|
| HLE | 44% | Scored | Extremely difficult questions contributed by domain experts worldwide to test frontier AI |
| AA-Omniscience Accuracy | 17% | Scored | An Artificial Analysis knowledge metric for the proportion of correctly answered questions. |
| AA-HLE | 11% | Scored | An Artificial Analysis Humanity's Last Exam score. |
| HLE w/o tools | 7% | Scored | Tool-free variant of Humanity's Last Exam used to isolate raw frontier reasoning without external aids |
| MMLU-Pro | 6% | Scored | Harder version of MMLU with 10 answer choices and more reasoning-focused questions |
| MMLU-Pro (Vals) | 6% | Scored | Vals AI’s independent run of MMLU-Pro across fourteen academic subjects. |
| GPQA | 3% | Scored | Expert-level questions in biology, physics, and chemistry |
| AA-GPQA Diamond | 2% | Scored | An Artificial Analysis GPQA Diamond score. |
| SuperGPQA | 2% | Scored | Enhanced version covering 285 disciplines |
| GPQA Diamond (Vals) | 2% | Scored | Vals AI’s independent run of the GPQA Diamond graduate-level science questions. |
| MMLU | — | Display only | Tests knowledge across 57 academic subjects |
| GPQA-D | — | Display only | Provider-table reference for GPQA Diamond scores reported in first-party comparison charts. |
| HLE-Diamond | — | Display only | The benchmark owner's high-reasoning, no-tools results on 1,000 refined HLE questions |
| HLE-Diamond (highest effort) | — | Display only | The benchmark owner's highest-available-effort, no-tools results on HLE-Diamond |
| HLE-Diamond (web+code) | — | Display only | The benchmark owner's high-reasoning HLE-Diamond results with restricted web and code tools |
| FrontierScience | — | Display only | Research-level science and scientific reasoning benchmark |
| SimpleQA | — | Display only | Factual question answering benchmark |
| HealthBench Hard | — | Display only | A harder health reasoning benchmark subset used in first-party frontier model comparisons. |
| HealthBench Professional | — | Display only | An open benchmark for clinician-facing model responses across care consult, writing and documentation, and medical research tasks. |
| MedXpertQA (Text) | — | Display only | Medical multiple-choice benchmark covering many specialties with text-only questions. |
| FrontierScience Research | — | Display only | A research-oriented FrontierScience variant focused on scientific investigation and solution quality. |
| MMLU-Pro (Arcee) | — | Display only | Display-only MMLU-Pro reference from Arcee AI's Trinity-Large-Thinking launch chart. |
| MMMLU | — | Display only | A multilingual MMLU-style benchmark reported in provider evaluation tables. |
About Knowledge benchmarks
Tests knowledge across 57 academic subjects
Questions
What is the best LLM for knowledge tasks?
The top LLMs for knowledge tasks are ranked by benchmarks like MMLU and GPQA, which test factual accuracy and expert-level understanding across dozens of subjects.
What is MMLU and how does it measure LLM knowledge?
MMLU (Massive Multitask Language Understanding) tests LLMs across 57 subjects from STEM to humanities, measuring broad factual knowledge and reasoning at varying difficulty levels.
What benchmarks test knowledge in AI models?
Key knowledge benchmarks include MMLU, MMLU-Pro, GPQA, SuperGPQA, HLE, and FrontierScience, each evaluating different depths of factual and scientific understanding.
How do knowledge benchmarks differ from reasoning benchmarks?
Knowledge benchmarks focus on factual recall and domain expertise, while reasoning benchmarks test logical deduction and multi-step problem solving independent of specific facts.
Knowledge benchmark updates
Know which model knows the most — updated every week.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.
Related
AI Web Scraping Tools
Compare tools that collect web data for retrieval and monitoring.
Web Data Pipeline
See how collection, monitoring, and validation work together.
Best LLMs Overall
Top models ranked across all benchmark categories.
Reasoning Benchmarks
Multi-step inference and logical deduction leaderboard.
Multilingual Benchmarks
Cross-language knowledge and reasoning scores.
LLM Selector Quiz
Find the best model for knowledge-intensive work.