Gemini API gemini-3.5-flash-lite; high thinking; default sampling; pass@1 except where noted.
未知项
Exact thinking budget and sampling parameters are not published.
工具
允许工具
推理预算
high
框架
未记录
协议
OSWorld-Verified official methodology; averaged over 5 runs, one attempt each, Docker/evaluator, 1080p, max 100 steps, pyautogui and UI-specific function declarations.
条件
Partial score; exact task set revision not stated.
Gemini 3.5 Flash-Lite · OSWorld-Verified · Partial score; 5 runs; 100-step maximum · Scores and configurations
Gemini 3.5 Flash-Lite record for OSWorld-Verified · Partial score; 5 runs; 100-step maximum. 1 score record for this release, including 1 exact value (may include historical records). Includes 1 source location.
Gemini 3.5 Flash-Lite · OSWorld-Verified scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Partial score74%Reported 74.0% · Unit %
Published Unconfirmed
Gemini 3.5 Flash-Lite high thinking
Settings
Gemini API gemini-3.5-flash-lite; high thinking; default sampling; pass@1 except where noted.
Unknowns
Exact thinking budget and sampling parameters are not published.
Tools
Tools allowed
Reasoning budget
high
Framework
not recorded
Protocol
OSWorld-Verified official methodology; averaged over 5 runs, one attempt each, Docker/evaluator, 1080p, max 100 steps, pyautogui and UI-specific function declarations.
Conditions
Partial score; exact task set revision not stated.