Compare AI models

Specifications
Parameters
—27B/35B/122B
Context window
1M256k
Benchmarks
Humanity's Last Exam · no tools
40.2%25.3%
OSWorld-Verified
78.4%58%
CharXiv Reasoning
84.2%77.2%
MMMU-Pro
83.6%76.9%
Overview
CompanyGoogleQwen
Release dateMay 19 2026Feb 24 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Gemini 3.5 Flash leads Qwen3.5 Medium on 4 of the 4 benchmarks they both report (Humanity's Last Exam, OSWorld-Verified, CharXiv Reasoning, MMMU-Pro). Qwen3.5 Medium shipped 84 days before Gemini 3.5 Flash, so benchmark comparisons should account for the intervening progress.

Context windows are 1M (Gemini 3.5 Flash) vs 256k (Qwen3.5 Medium). Gemini 3.5 Flash is closed, while Qwen3.5 Medium is open weight.

On Humanity's Last Exam · no tools, Gemini 3.5 Flash leads at 40.2% vs Qwen3.5 Medium at 25.3%. On OSWorld-Verified, Gemini 3.5 Flash leads at 78.4% vs Qwen3.5 Medium at 58%. On CharXiv Reasoning, Gemini 3.5 Flash leads at 84.2% vs Qwen3.5 Medium at 77.2%. On MMMU-Pro, Gemini 3.5 Flash leads at 83.6% vs Qwen3.5 Medium at 76.9%.