模型支持企业博客
500+ AI 模型 API,一次搞定,就在 CometAPI
模型 API
开发者
快速入门文档API 仪表板
资源
AI 模型博客企业更新日志关于
2025 CometAPI。保留所有权利。隐私政策服务条款
Home/Models/Google/Gemini 2.5 Pro
G

Gemini 2.5 Pro

输入:$1/M
输出:$8/M
上下文:1M
最大输出:65K
Gemini 2.5 Pro 是 Google 提供的人工智能模型。它具备原生多模态处理能力和高达 1 million tokens 的超长上下文窗口,为复杂、长序列任务提供前所未有的强大支持。根据 Google 的数据,Gemini 2.5 Pro 在复杂任务上表现尤为出色。该模型支持的最大上下文长度为 1,048,576 tokens。
新
商用
Playground
概览
功能亮点
定价
API
版本

Basic Information (Features)

  • Multimodality: Natively handles text, images, and code in a single model.
  • Long Context Window: Maintains coherence over extended conversations and documents.1.05M
  • Deep Think Mode: An experimental variant within the Pro suite that deploys multiple reasoning agents in parallel for strategic planning and creative solutions.
  • Ideal Use Cases: Coding, agentic workflows, interactive simulations, and data visualization .

Technical Details

  • Multi-Agent Architecture: Parallelizes reasoning streams to explore multiple solution paths simultaneously.
  • MRCR (Multi-Round Coreference Resolution): Enhanced co-reference handling for sustained dialogues and multi-turn tasks.
  • Training Corpus: Billions of tokens spanning web text, code repositories, academic sources, and proprietary datasets.
  • Tool Integration: Seamlessly combines code execution, Google Search, and external APIs to augment its internal reasoning.
  • Limitations & known risks
  • Content policy constraints: models enforce content policies (e.g., disallowing explicit sexual content and some illicit content), but enforcement is not perfect — generating images of public figures or controversial icons may still be possible in some scenarios, so policy checks are essential. )
  • Failure modes: possible identity drift in extreme edits, occasional semantic misalignment (when prompts are under-specified), and artifacts in very complex scenes or extreme viewpoint changes.
  • Provenance & misuse: while watermarks and SynthID are present, these do not prevent misuse — they assist detection and attribution but are not a substitute for human review in sensitive workflows.

Typical use cases

  • Product & ecommerce: place/catalog products into lifestyle shots via multi-image fusion.
  • Creative tooling / design: fast iterations in design apps (Adobe Firefly integration cited).
  • Photo editing & retouching: localized edits from natural language (remove objects, change color/lighting, restyle).
  • Storytelling / character assets: keep characters consistent across panels and scenes.

Gemini 2.5 Pro 的功能

了解 Gemini 2.5 Pro 的核心能力,帮助提升性能与可用性,并改善整体体验。

Gemini 2.5 Pro 的定价

查看 Gemini 2.5 Pro 的竞争性定价,满足不同预算与使用需求,灵活方案确保随需求扩展。

gemini-2.5-pro (same price across variants)

Model familyVariant (model name)Input price (USD / 1M tokens)Output price (USD / 1M tokens)
gemini-2.5-progemini-2.5-pro-all$1.00$8.00
gemini-2.5-progemini-2.5-pro-thinking$1.00$8.00
gemini-2.5-progemini-2.5-pro$1.00$8.00

Gemini 2.5 Pro 的示例代码与 API

获取完整示例代码与 API 资源,简化 Gemini 2.5 Pro 的集成流程,我们提供逐步指导,助你发挥模型潜能。
POST
/v1beta/models/{model}:{operator}
POST
/v1/chat/completions
Python
JavaScript
Curl
from google import genai
import os

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com"

client = genai.Client(
    http_options={"api_version": "v1beta", "base_url": BASE_URL},
    api_key=COMETAPI_KEY,
)

response = client.models.generate_content(
    model="gemini-2.5-pro",
    contents="Tell me a three sentence bedtime story about a unicorn.",
)

print(response.text)

Python Code Example

from google import genai
import os

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com"

client = genai.Client(
    http_options={"api_version": "v1beta", "base_url": BASE_URL},
    api_key=COMETAPI_KEY,
)

response = client.models.generate_content(
    model="gemini-2.5-pro",
    contents="Tell me a three sentence bedtime story about a unicorn.",
)

print(response.text)

JavaScript Code Example

import { GoogleGenAI } from "@google/genai";

// Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
const COMETAPI_KEY = process.env.COMETAPI_KEY || "<YOUR_COMETAPI_KEY>";
const BASE_URL = "https://api.cometapi.com";

async function main() {
  const ai = new GoogleGenAI({
    apiKey: COMETAPI_KEY,
    httpOptions: { baseUrl: BASE_URL },
  });

  const response = await ai.models.generateContent({
    model: "gemini-2.5-pro",
    contents: "Tell me a three sentence bedtime story about a unicorn.",
  });

  console.log(response.text);
}

main();

Curl Code Example

from google import genai
import os

# Get your CometAPI key from https://api.cometapi.com/console/token, and paste it here
COMETAPI_KEY = os.environ.get("COMETAPI_KEY") or "<YOUR_COMETAPI_KEY>"
BASE_URL = "https://api.cometapi.com"

client = genai.Client(
    http_options={"api_version": "v1beta", "base_url": BASE_URL},
    api_key=COMETAPI_KEY,
)

response = client.models.generate_content(
    model="gemini-2.5-pro",
    contents="Tell me a three sentence bedtime story about a unicorn.",
)

print(response.text)

Gemini 2.5 Pro 的版本

Gemini 2.5 Pro 可能存在多个快照,原因包括:更新后保持一致性需要保留旧版、给开发者留出迁移窗口,以及全球/区域端点提供的优化差异。具体差异请参考官方文档。
gemini-2.5-pro-all
gemini-2.5-pro-thinking
gemini-2.5-pro-deepsearch

更多模型

A

Claude Opus 4.6

输入:$4/M
输出:$20/M
Claude Opus 4.6 是 Anthropic 的“Opus”级大型语言模型,于 2026 年 2 月发布。其定位为知识工作与研究工作流的主力模型——提升长上下文推理、多步骤规划、工具使用(包括代理型软件工作流),以及计算机使用类任务,如自动生成幻灯片和电子表格。
A

Claude Sonnet 4.6

输入:$2.4/M
输出:$12/M
Claude Sonnet 4.6 是迄今为止我们最强大的 Sonnet 模型。它对模型在编码、计算机使用、长上下文推理、智能体规划、知识工作和设计等方面的能力进行了全面升级。Sonnet 4.6 还在 beta 阶段提供 1M token 上下文窗口。
O

GPT-5.4 nano

输入:$0.16/M
输出:$1/M
GPT-5.4 nano 专为速度和成本最为关键的任务而设计,例如分类、数据提取、排序以及子智能体。
O

GPT-5.4 mini

输入:$0.6/M
输出:$3.6/M
GPT-5.4 mini 将 GPT-5.4 的优势融入到一款更快速、更高效、专为大规模工作负载设计的模型中。
A

Claude Mythos Preview

A

Claude Mythos Preview

即将推出
输入:$60/M
输出:$240/M
Claude Mythos Preview 是我们迄今为止最强大的前沿模型,并显示出 在许多评测基准上的得分相较于我们此前的前沿模型 Claude Opus 4.6 有显著跃升。
X

mimo-v2-pro

输入:$0.8/M
输出:$2.4/M
MiMo-V2-Pro 是 Xiaomi 的旗舰基础模型,拥有超过 1T 的总参数量和 1M 的上下文长度,并针对智能体场景进行了深度优化。它对 OpenClaw 等通用智能体框架具有很强的适配性。在标准 PinchBench 和 ClawBench 基准测试中,它跻身全球第一梯队,感知性能接近 Opus 4.6。MiMo-V2-Pro 旨在作为智能体系统的大脑,协调复杂工作流,推动生产工程任务,并可靠地交付结果。

相关博客

免费版 Gemini 2.5 Pro API 是否出现故障?2025 年免费配额的变更
Dec 11, 2025
gemini-2-5-pro
gemini-2-5-flash

免费版 Gemini 2.5 Pro API 是否出现故障?2025 年免费配额的变更

Google 已大幅收紧 Gemini API 的免费层:Gemini 2.5 Pro 已从免费层移除,Gemini 2.5 Flash 的每日免费请求被大幅削减(据报道:~250 → ~20/天)。这并不意味着该模型在实验用途上已永久“死掉”,但这确实意味着对于许多现实世界的用例,免费访问实际上已被大幅削弱。
Gemini 在所有层级的使用限额有哪些限制?
Sep 26, 2025
gemini-2-5-pro
veo-3

Gemini 在所有层级的使用限额有哪些限制?

Google 已将模糊的“有限访问”措辞改为针对 Gemini 应用(免费、Google AI Pro 和 Google AI Ultra)的明确、按层级划分的上限。这些上限