Gemini 1.5 Pro: benchmark results

Google's MoE-based Gemini 1.5 flagship, notable for long-context multimodal understanding of up to 2M tokens (February 2024). Provider: Google. Released 2024-02-15. Access: API.

Unified ELO 1556 ± 1, rank #416 of 1392 rated models, from 611 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AV-Odyssey - Music Score Matching29Score (%)100
AV-Odyssey - Speech Sentiment Analysis29.5Score (%)100
Eval Anything4.58Overall score100
Loong55.37Overall Avg Score (%)100
MMNeedle (1 Image, 8*8 Stitching, Exact Accuracy)29.811 Image, 8*8 Stitching, Exact Accuracy100
OpenVLM MMT-Bench - Handwritten Retrieval95Score (%)100
Vector Eval - IFEval89.82Final Accuracy (%)100
OpenVLM MMT-Bench - Texture Material Recognition90Score (%)99.5
TextClass Benchmark1782.7Meta-Elo (self-reported)99.1
MERA - PARus95.8Accuracy (%)99
OpenVLM MMT-Bench - 3D CAD Recognition80Score (%)98.5
HELMET (128K)64.4Average Score98.4

Interactive version: theaggregate.ai/model?slug=gemini-1-5-pro · How It Works · Data refreshed daily, snapshot 2026-09-05.