AITOPAI热报

Gemini 3.1 Pro

Google·US·Released 2026-02-19
ClosedPreviewGeneral LLMAudioReasoningTool useVision
1Mcontext
$2 / $12per 1M tokens (in / out)
—parameters
11stories in our feed

Specs

hover a value to see its source
Context window1M
Max output64K
Parameters—
Knowledge cutoff2025-01
LicenceProprietary
Commercial useAPI only
InputText · Image · Video · Audio · PDF
OutputText
Familygemini-pro

API price

USD per 1M tokens
Input$2
Output$12
Cached input$0.2

No price change recorded since we started tracking.

News timeline

from the AITOP feed, one story per eventAll 11 →
2026-10-28
Google DeepMind 与 Stanford 研究：图片符号能隐秘操纵模型写递归代码DAIR.AI
2026-10-23
V4.1 Flash-1017 在 ARC-AGI-2 上略超 Gemini 3.1 Pro PreviewTeortaxes
2026-08-05
LLM评审与人类同行评审对齐程度如何arXiv: OpenAI
2026-07-15
Dilum Sanjaya 开始新系列：用 AI 搭建未来感 UIAI Will
2026-07-14
沙漏推理：通过结构隔离增强少样本归纳推理arXiv cs.AI
2026-07-13
用GPT Images 2 + Gemini 3.1 Pro制作Vibe Coding风格游戏UI工作流AI Will
2026-07-03
Rubric评估对比GPT 5.4、Claude Opus 4.7、Gemini 3.1 Pro临床推理能力arXiv cs.LG
2026-06-27
DeepSeek Pro Max 模式在编码基准上超越 GPT-5.4、Gemini 3.1 Pro 等Geek
2026-06-25
《华盛顿邮报》调查：主流AI聊天机器人政治立场左倾，自称'反觉醒'的Grok也不例外Decoder