AI IQ
Subscribe

Reliability

A bird's-eye view of the dimension: the main IQ chart, the effective-cost frontier, and the benchmark charts that feed it.

Reliability IQ
Reliability IQ
Each model's Reliability IQ plotted on a standard normal IQ distribution
Controls:

How to read this chart

Models are placed on a standard IQ-style distribution using Reliability IQ. Higher scores appear farther to the right.

Benchmarks in this dimension

AA Long Chain ReasoningAA OmniscienceIFBenchBullshitBench v2FACTS Grounding
Reliability vs Effective Cost
Reliability vs Effective Cost
X = effective cost (log). Y = Reliability IQ. Color = provider.
Controls:
Reliability1:1Cost

How to read this chart

Each point is a public model. The chart compares Reliability IQ against Effective Cost (per 1M I/O Tokens), with color showing the model provider.

Data sources
Artificial AnalysisARC PrizeVals.aiBullshitBench v2Kaggle FACTS Grounding leaderboard
Benchmarks In This Dimension
AA Long Chain Reasoning Scores
Artificial Analysis Long Chain Reasoning scores for long-document extraction, synthesis, and reasoning. Color = provider.
Controls:

How to read this chart

Bars rank models by the source-backed benchmark value used for this chart. Longer bars indicate higher published scores.

Data sources
Artificial Analysis
AA Omniscience Scores
Artificial Analysis Omniscience scores for factual reliability. Color = provider.
Controls:

How to read this chart

Bars rank models by the source-backed benchmark value used for this chart. Longer bars indicate higher published scores.

Data sources
Artificial Analysis
IFBench Benchmark Scores
Instruction-following and constraint-adherence scores. Color = provider.
Controls:

How to read this chart

Bars rank models by the source-backed benchmark value used for this chart. Longer bars indicate higher published scores.

Data sources
Artificial Analysis
BullshitBench v2 Scores
Clear Pushback rate: share of attempts where a model clearly challenges a false premise instead of accepting nonsense. Color = provider.
Controls:

How to read this chart

Bars rank models by the source-backed benchmark value used for this chart. Longer bars indicate higher published scores.

Data sources
BullshitBench v2
FACTS Grounding Scores
Google FACTS Grounding scores for long-form responses fully grounded in provided documents. Color = provider.
Controls:

How to read this chart

Bars rank models by the source-backed benchmark value used for this chart. Longer bars indicate higher published scores.

Data sources
Kaggle FACTS Grounding leaderboard