OpenAI
GPT-Image 2
OpenAI’s second-generation image model — instruction-precise synthesis, editing and in-image text.
智商指数
93.1 #8
AAQI 93.0
输出速度
25 t/s
ITL 40.0 ms/tok
混合价格 (3:1)
$0.09 /1M
输入 $0.09/M · 输出 $0.09/M
上下文
128K
≈ 320 页 A4 · 最大输出 4,096
Quality Domain Radar
6-Axis Frontier Evaluation vs Category Average & Flagship
GPT-Image 2
Claude Mythos Preview (#1)
Average (68)
缺少直接基准测量时,雷达图各轴由类别得分近似得出 — 详见 评分方法。
优势与取舍
优势所在
- Best instruction-following in the image class
- Clean typography and UI-style text rendering
- First-class edit/inpaint endpoints
- Deep integration with OpenAI agents
需要注意的权衡
- Most expensive per-image tier
- Slower renders on complex scenes
- Closed weights
杀手级用例
Design-to-asset pipelines
Prompt-precise banners, mockups and brand assets.
Agentic creative tools
Image calls orchestrated by GPT-6 agents.
Text-heavy graphics
Posters and slides needing real, legible copy.
开发者快速上手与 API
请先设置
OPENAI_API_KEY — 获取 API 密钥curl https://api.openai.com/v1/responses \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-H "content-type: application/json" \
-d '{
"model": "gpt-image-2",
"input": "Refactor this function for O(n) time and explain the change."
}'from openai import OpenAI
client = OpenAI() # reads OPENAI_API_KEY
resp = client.responses.create(
model="gpt-image-2",
input="Refactor this function for O(n) time and explain the change.",
)
print(resp.output_text)const res = await fetch("https://api.openai.com/v1/responses", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.OPENAI_API_KEY}`,
"content-type": "application/json",
},
body: JSON.stringify({ model: "gpt-image-2", input: "Refactor this function for O(n) time and explain the change." }),
});
console.log((await res.json()).output_text);官方资源与链接
API 部署
对比托管商与延迟
Groq、Cerebras、Together、DeepInfra、Fireworks、Bedrock 与 Azure 的吞吐量、TTFT 和缓存定价。
GroqCerebrasTogether AIDeepInfraFireworks AIAWS BedrockAzure AI
各榜单得分
分类能力得分
基准测试分项
实时对决
实时规格与架构
底层解析
GPT-Image 2 — OpenAI image generation
- Second-gen image stack trained with GPT-side prompt understanding
- Legible in-image text and dense layout control
- Transparent-background and masked-edit modes
- Tight GPT-6/Responses API integration
发布日期
2026年10月16日
架构
GPT-Image 2 — OpenAI image generation
参数量
Not disclosed
开源协议
Proprietary
上下文窗口
128,000 tokens
最大输出
4,096 tokens
模态
text + image
缓存价格
读 $0.02 · 写 $0.11 /1M
推理
—
输入价格
$0.09/M
输出价格
$0.09/M
输出速度
25 t/s
首Token延迟
—
数据更新
2026年10月22日
每月账单预估
每月预估支出
$0.00
正在计算相对 GPT-4o 的节省额…
引用与嵌入
GitHub Markdown 徽章
Loading badge markdown...BibTeX
@misc{llmpodium2026gptimage2,
title = {Benchmark Analysis and Intelligence Evaluation of GPT-Image 2},
author = {LLMPodium Research},
year = {2026},
url = {https://llmpodium.com/models/gpt-image-2}
}嵌入评分徽章
将实时 Podium 评分徽章添加到 GitHub 或 Hugging Face README —— 链接回实时排名。
预览
Markdown
Loading badge snippet...HTML
Loading badge snippet...更多 OpenAI 模型
替代模型推荐
基于能力、定价和基准测试得分与 GPT-Image 2 最接近的替代模型。