Atlas

Models

← All models

Gemini 1.5 Flash (source unspecified)

Google

This benchmark entry preserves generative-model observations labeled Gemini 1.5 Flash or an unversioned provider alias when the source does not identify an exact checkpoint. Google released distinct preview, 001, and 002 versions and explicitly redirected the gemini-1.5-flash alias to 002 in November 2024. The stable Atlas slug is retained for unresolved observations; neither an evaluation date nor a May 2024 family label alone establishes an exact API snapshot.

Benchmark scores

BenchmarkScore
A-Fantasia74.7
A-Fantasia (Valid Attempts with Retries)74.7
A-Fantasia Backwards Spelling100.0
A-Fantasia Chess58.0
A-Fantasia Chess (Valid Attempts with Retries)58.0
A-Fantasia Cube (Valid Attempts with Retries)66.0
A-Fantasia Cube Rotation66.0
A-Fantasia Spelling (Valid Attempts with Retries)100.0
AI2D (OpenVLM)78.5
AidanBench1014
Artificial Analysis Progress Index v1139
Epoch Capabilities Index (ECI)123
GeoBench3934
GeoBench ACW Country Accuracy76.0
GeoBench Photos3458
GeoBench Photos Country Accuracy64.0
GPQA Diamond (Artificial Analysis avg@5)32.4
HallusionBench (OpenVLM)48.5
HardGeoBench <=100mi32.5
HardGeoBench <=10mi16.3
HardGeoBench Avg Distance1763
HardGeoBench Country Accuracy48.8
Humanity's Last Exam (Text-Only)4.4
LiveCodeBench (Artificial Analysis)19.6
LLM Deception Disinformation Effectiveness0.6
LLM Deception Vulnerability Score0.6
LLM Divergent Thinking Repeat Rate0.7
LLM Divergent Thinking Score4.1
MATH-500 (Artificial Analysis source)55.4
MathVista (OpenVLM)51.3
Loading Atlas data…