Unverified50% confidenceBenchmarkExact time
在中性模式下，Doubao Seed 2.0 Pro端到端准确率最高，为43.1%，Gemini 3.1 Pro Preview为40.8%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/9/2026
First Seen
Valid until: 10/7/2026
Sources
腾讯DiscoBench：评测LLM搜索代理歧义澄清能力的首个基准
bilibili熊二等兵7/4/2026
Related Claims
Unverified在中性模式下，Gemini 3.1 Pro Preview端到端准确率为40.8%78% similarUnverifiedGemini 3.5 Pro在Video ModeM多模态视频理解基准中得分82.1分，大幅领先其他模型70% similarUnverifiedGemini 3.1 Pro High在SWE-bench测试中通过率为84.17%，排名第一，生成代码量为307,000行，Bug密度为614个/百万行代码，安全问题为210个/百万行代码69% similarUnverifiedGemini 3.5 Flash的编码能力基准测试得分为76.2%68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/387220API
curl https://kongchang.com/api/v1/knowledge/claims/387220MCP
get_claim(id=387220)