UnifyBench

Weighted consensus · Public benchmarks

LLM leaderboard

Compare AI models across public benchmarks. Source-balanced consensus for overall performance, coding, reasoning and agentic work.

Looking for which LLM is best? UnifyBench is a first stop: source-balanced consensus across many published benchmarks (not one board). Shortlist here, then validate on Artificial Analysis, check Arena for preference, and finish with task-specific evals.

See the full ranking

Overall leaders

Comparison score · 0–100

Consensus across available benchmarks. Models cover different task areas; use a task ranking for a focused comparison.

Experimental scores for the selected profile and sources. Small gaps may not be meaningful. How scoring works · Effort evidence collected

460unique models545model configurations127source pages
Explore all 127 source pages Independent evaluations and lab reports

Each model is counted once across effort levels. Configurations count documented model + effort settings, including unranked.

Original reports and benchmark leaderboards. Distinct source pages with published results in the catalog. Multiple results and page sections count as one source.

One entry per model, using its highest-scoring eligible effort setting. How effort ranking works.

Customize ranking Sources, profiles and evidence details
All presets

Evidence only MultimodalLong contextHuman prefAll capabilitiesThese profiles do not yet have enough comparable evidence for a ranking.

Adjust weights

Full ranking

223 entries
Compare two configurations

Capability ranking of measured model configurations. Scores are experimental; missing results remain unknown.
RankModelScoreEvidence breadth Why this score?
1GPT-6 Astra · XHighOpenAI95.9Multiple independent sources21 families · 6 areas measured
2Claude Opus 5.5 · HighAnthropic92.4Multiple independent sources14 families · 6 areas measured
3Claude Fable 5.1 · MaxAnthropic91.5Multiple independent sources37 families · 6 areas measured
4GPT-6.1 Sol · HighOpenAI86.0Multiple independent sources16 families · 6 areas measured
5Claude Opus 5 · MediumAnthropic83.7Multiple independent sources14 families · 6 areas measured
6Claude Sonnet 5.5 · XHighAnthropic81.4Multiple independent sources19 families · 5 areas measured
7Gemini 4 Argon · HighGoogle80.3Multiple independent sources16 families · 5 areas measured
8Claude Fable 5 · HighAnthropic80.0Multiple independent sources9 families · 2 areas measured
9GPT-5.6 Sol · XHighOpenAI78.9Multiple independent sources17 families · 6 areas measured
10GPT-5.5 Pro · XHighOpenAI74.8Multiple independent sources3 families · Hard reasoning only
11Muse Spark 1.3 · XHighMeta74.8Multiple independent sources28 families · 6 areas measured
12Gemini 3.8 Flash · HighGoogle74.5Multiple independent sources37 families · 6 areas measured
13Kimi K3 · MaxMoonshot67.2Multiple independent sources36 families · 6 areas measured
14Grok 4.6 · MediumxAI64.2Multiple independent sources14 families · 5 areas measured
15GPT-6 Sol · MaxOpenAI63.6Multiple independent sources35 families · 6 areas measured
16Gemini 3.7 Flash · MediumGoogle63.0Multiple independent sources13 families · 6 areas measured

More models on this page

Behind the ranking

284 entries with published results · 223 ranked · 7 provisional estimates.

Explore seven capabilities, the original results, and gaps in the evidence.

Coding rankingReasoning rankingAgentic rankingKnowledge rankingCompare capabilities ↗Published benchmarks ↗Evidence coverage ↗
Cite this ranking

Include the model and effort setting, the selected profile, and the dated data downloads. A shared link uses the latest data; it does not freeze a ranking in time.

UnifyBench ranking
Data updated: 6 Oct 2026
Mode: overall; evidence: documented; configurations: best
https://unifybench.ai/

Catalog data · Effort measurements · Configuration scoring rules · Methodology

Catalog dated 6 Oct 2026. Effort evidence collected 6 Oct 2026. Collection dates are not evaluation dates.

What does the score mean?

The adjusted comparison score estimates performance against the same reference panel from matched published comparisons. Capability pools matched comparisons across families and checks aggregate breadth and independent source coverage. Custom profiles require support in every selected capability. Missing results never count as losses.

Family details show observed win shares against available reference peers. Those descriptive values are supporting evidence, not adjusted capability scores. Provider reports and independent results remain labeled. Matching reported settings do not establish equal inference effort, and score differences do not establish statistical significance.

Read the full methodology →