Benchmark Radar™
RSS Contact Star —

AGI Eval

Explore score history ↗

AGI Eval is an AI benchmark in the Benchmark Radar catalog.

Released
2023-04-14
Openness
unknown
Importer
Benchmark Saturation

Source provenance

AGI Eval results and reported scores

Reported benchmark scores · Benchmark Saturation

Reported scores: 5

ModelOrganizationReported valueReportedEvidence
Deepseek v3 base—79.6—evidence
GPT-4o—71.4—evidence
Kimi k2—84.23—evidence
Llama 3 400B+—69.9—evidence
Olmo 3—88.8—evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.

Which sources cite AGI Eval?