Unverified50% confidenceFactExact time
Google Gemini 使用 contents 字段，角色标识为 user 和 model，不支持 system 角色直接注入，而是通过 systemInstruction 字段实现
1
Sources
50%
Confidence
Long-term
Relevance
7/7/2026
First Seen
Sources
Kontext：一键迁移AI对话上下文，打破多模型切换壁垒
hackernewshackernews7/3/2026
Related Claims
UnverifiedUsers can choose to authorize their Google apps to Gemini in the settings, and once connected, Gemini can automatically fill in image generation details that are not explicitly specified.81% similarUnverifiedGoogle's Gemini model family has already been deployed on-device on Android devices, supporting capabilities such as real-time translation, intelligent summarization, and multimodal understanding.76% similarUnverified在Google Gemini API的响应结构中，finishReason字段用于指示模型为何停止生成内容75% similarUnverifiedGoogle's choice to create Neural Expressive for Gemini rather than continuing with Material Design signifies that Google is operating Gemini as an independent brand.75% similarUnverifiedGemini's personalized image generation feature understands individual preferences and generates visual content by connecting through the user's Google app ecosystem.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/195026API
curl https://kongchang.com/api/v1/knowledge/claims/195026MCP
get_claim(id=195026)