Gemini 3 Pro 的 SWE-bench Verified · Verified subset (500 problems) 官方出处。Gemini 3 Pro API default · 76.2%。见《A new era of intelligence with Gemini 3》Release article / Build anything paragraph。证据 main-swe-bench。
Gemini 3 Pro · SWE-bench Verified(Verified subset (500 problems)),Gemini 3 Pro API default配置,Pass rate 76.2%(具体成绩公布时间未确认)。见《A new era of intelligence with Gemini 3》Release article / Build anything paragraph。 查看成绩条件
Article page collected 2026-11-10; article date 2025-11-18
版本公布
披露时间
未确认 · 首发/后补未确认
采集时间
历史依据
暂无可靠历史版本依据
引用文本
Gemini 3 Pro · SWE-bench Verified(Verified subset (500 problems)),Gemini 3 Pro API default配置,Pass rate 76.2%(具体成绩公布时间未确认)。官方出处 A new era of intelligence with Gemini 3(主文,Article page collected 2026-11-10; article date 2025-11-18)Release article / Build anything paragraph,https://blog.google/products-and-platforms/products/gemini/gemini-3/。Benchlog https://benchlog.lekshome.cc/evidence/gemini-3-pro/main-swe-bench/
关联成绩与条件
Gemini 3 Pro · SWE-bench Verified 成绩与配置
指标与成绩
配置与评测条件
证据关系
Pass rate76.2%原文 76.2% · 单位 %
公布 未确认
Gemini 3 Pro API default
设置
Gemini API model id gemini-3-pro-preview; default sampling settings. Gemini results are pass@1 and single-attempt unless specified; smaller benchmarks average multiple trials.
未知项
The evaluation table does not specify a thinking budget or inference-time compute setting.
工具
允许工具
推理预算
未记录
框架
未记录
协议
Single attempt; average over 10 runs; bash, file-operation and submit tools.
条件
Official provider methodology; different scaffolds and infrastructure mean results are not directly comparable.
Gemini 3 Pro · SWE-bench Verified (Verified subset (500 problems)) 76.2% · Main text evidence (main-swe-bench)
Official source for SWE-bench Verified · Verified subset (500 problems) in Gemini 3 Pro. Gemini 3 Pro API default · 76.2%. See “A new era of intelligence with Gemini 3” at Release article / Build anything paragraph. Evidence main-swe-bench.
Gemini 3 Pro · SWE-bench Verified (Verified subset (500 problems)), configuration Gemini 3 Pro API default, Pass rate 76.2% (exact score publication date unconfirmed). See “A new era of intelligence with Gemini 3” at Release article / Build anything paragraph. View score conditions
Main text · Release article / Build anything paragraph
Material version
Article page collected 2026-11-10; article date 2025-11-18
Version published
Disclosure date
Unconfirmed · Launch or later disclosure unconfirmed
Collected
Historical basis
No reliable historical version evidence available
Citation text
Gemini 3 Pro · SWE-bench Verified (Verified subset (500 problems)), configuration Gemini 3 Pro API default, Pass rate 76.2% (exact score publication date unconfirmed). Official source: A new era of intelligence with Gemini 3 (Main text, Article page collected 2026-11-10; article date 2025-11-18), Release article / Build anything paragraph, https://blog.google/products-and-platforms/products/gemini/gemini-3/. Benchlog https://benchlog.lekshome.cc/evidence/gemini-3-pro/main-swe-bench/?lang=en
Related scores and conditions
Gemini 3 Pro · SWE-bench Verified scores and configurations
Metric and score
Configuration and evaluation conditions
Evidence relation
Pass rate76.2%Reported 76.2% · Unit %
Published Unconfirmed
Gemini 3 Pro API default
Settings
Gemini API model id gemini-3-pro-preview; default sampling settings. Gemini results are pass@1 and single-attempt unless specified; smaller benchmarks average multiple trials.
Unknowns
The evaluation table does not specify a thinking budget or inference-time compute setting.
Tools
Tools allowed
Reasoning budget
not recorded
Framework
not recorded
Protocol
Single attempt; average over 10 runs; bash, file-operation and submit tools.
Conditions
Official provider methodology; different scaffolds and infrastructure mean results are not directly comparable.