The Scale
Comparing
BRAIAIN Standard
The mathematical portrait of every AI model. Scored 400–1600 from 6 million human votes.

Every model has
a mathematical
portrait.

Eight dimensions of capability. One Lissajous curve. No two models produce the same figure. Derived from Chatbot Arena — 6M+ crowdsourced votes.

Rankings
Loading…
Scores are derived from Chatbot Arena Elo ratings. Arena ratings are based on 6 million+ blind pairwise human votes across 347 models at arena.ai. BRAIAIN normalizes Arena category Elo scores across 8 dimensions to produce a single 400–1600 composite and a unique DNA visualization per model. Different tasks, prompts, and real-world conditions may produce different relative outcomes.

Data attribution: Evaluation data from Chatbot Arena by Arena Intelligence (lmarena-ai). BRAIAIN has no financial relationship with any model provider and receives no compensation tied to scores.

The unreached ceiling

When all eight dimensions reach 1.0, the figure achieves 8-fold rotational symmetry. The void at centre is the fixed point of dihedral group D₈. No model has earned it. The distance between the current leader and this figure is the entire story of where AI is right now.

BRAIAIN score
Analytical
Technical
Dimension profile

Scores derived from Chatbot Arena Elo ratings (6M+ votes).

Full model page ↗
What these results show: Every question asked, the model’s answer, and how it scored. BRAIAIN Standard measures cognitive capability across 8 dimensions. A high score does not indicate safety, alignment, fairness, or fitness for any specific use case. These are capability results only.

The BRAIAIN
Standard

Capability benchmark — scope statement
BRAIAIN Standard measures eight dimensions of cognitive capability. A high BRAIAIN score does not indicate safety, alignment, fairness, or fitness for any specific use case. These results measure what models can do, not whether they should be deployed, trusted, or preferred for your application.
What this is

BRAIAIN Standard visualizes AI model capability as mathematical art. We transform multi-dimensional performance data into unique Lissajous curves — one per model, derived from its scores across reasoning, math, coding, science, context, speed, efficiency, and multimodal capability.

Dimension scores are sourced from Chatbot Arena (arena.ai) — the industry-standard crowdsourced evaluation platform with 6 million+ human votes across 347 models. Arena Elo ratings across 8 categories are normalized to 0–1 to drive our scoring formula and DNA visualizations. We credit the Arena community for the underlying evaluation data.

BRAIAIN Standard has never accepted funding from any AI laboratory. The DNA visualization algorithm, print production, and editorial framing are ours. The evaluation data belongs to the community that generated it.

Scoring formula

Eight dimension scores (each 0.0–1.0) feed into two section scores (200–800 each) for a total of 400–1600.

Analytical (200–800):
  = 200 + (Reasoning×0.35 + Math×0.40 + Science×0.25) × 600

Technical (200–800):
  = 200 + (Coding×0.40 + Context×0.30 + Efficiency×0.15 + Speed×0.15) × 600

BRAIAIN Score = Analytical + Technical  [range: 400–1600]

Dimension score = Arena Elo normalized to 0.0–1.0
  (Elo 1350 = 0.0, Elo 1575 = 1.0)
Source: arena.ai — 6M+ crowdsourced votes
SectionMeasuresWeight
Analytical (200–800)Reasoning, mathematics, science50%
Technical (200–800)Coding, context, efficiency, speed50%
Independence policy

“BRAIAIN Standard has never accepted and will never accept funding, sponsorship, or paid placement from any AI laboratory or model provider. Scores are derived from public crowdsourced evaluation data (Chatbot Arena). Models are included at our discretion. The DNA visualization, scoring formula, and print production are independent of all model providers. If Arena ratings change, our scores update accordingly.”

How scoring works

Eight dimensions are drawn from Arena category leaderboards, each representing a distinct axis of model capability as evaluated by millions of human voters in blind pairwise comparisons.

BRAIAIN DimArena SourceWhat it measures
ReasoningHard PromptsComplex multi-step problem solving
MathMathMathematical reasoning and computation
CodingWebDev / CodeSoftware engineering and debugging
ScienceOverallGeneral knowledge and scientific reasoning
ContextMulti-TurnSustained coherence across exchanges
SpeedCreative WritingFluent, high-quality generation
EfficiencyInstruction FollowingPrecision in following constraints
MultimodalVision ArenaImage understanding and reasoning

Elo ratings are updated as new Arena votes are cast. BRAIAIN refreshes scores periodically. Raw Arena Elo values are preserved in our data for full transparency.

Score updates

BRAIAIN scores are derived from public Arena data and update accordingly.

  1. When Chatbot Arena publishes updated Elo ratings, BRAIAIN refreshes affected models
  2. New models are added when they appear in Arena with sufficient vote counts
  3. If you believe our normalization of Arena data contains an error, email hello@braiain.com
  4. Raw Arena Elo values are published alongside normalized BRAIAIN scores for verification
The DNA fingerprint

Each model’s eight dimension scores drive 16 harmonic components in a Lissajous parametric curve — all eight dimensions on both axes, but with prime frequencies (1, 2, 3, 5, 7, 11, 13, 17) on one axis and composite frequencies (1, 4, 6, 8, 9, 10, 14, 15) on the other. Each dimension’s score also sets the phase offset of another component, creating cross-coupling that ensures models with different capability profiles produce genuinely different shapes. A model strong in math but weak in coding traces a fundamentally different path than one strong in coding but weak in math, even at the same total score.

For educators

BRAIAIN Standard is used in AI literacy curricula.

Classroom resources, lesson plans, and bulk print licensing → educators.html
Submit a model

Any publicly available model may be submitted. No fee.

submissions@braiain.com

Own the
intelligence.

These figures exist only as long as these scores hold. When a model is retested and its score changes, the figure changes.

Each model’s benchmark scores produce a unique Lissajous figure — iterated 70,000 times, rendered as a scientific etching on archival cotton rag. Looks better in person than on screen.

Featured — Series print

The Scale

The complete arc of machine intelligence. Scores 400 through 1600 in sequence on one archival panel.

The floor. The ceiling. Every step between. One wall. One argument.

$149
24×36 in  ·  archival giclée  ·  signed

The Complete Argument

The Floor (400) + Current leader + Perfect score (1600) + The Scale. Four prints, one wall, complete story.

$199
Individual model prints
Archival giclée

Pigment-based inks rated for 100+ years of display lightfastness. Museum-quality reproduction process. These prints will outlast the models they depict.

Cotton rag paper

Acid-free, 100% cotton fibre. Warm white surface that complements the warm monochrome palette. The print looks better in person than on screen — the warm palette on cotton catches light in a way screens cannot reproduce.

Damaged in transit?

Printful replaces damaged prints at no cost. Contact us within 14 days with a photograph. No questions asked.