A free AI image generation tool. User provides a text description, and the agent generates a corresponding image. No API key required, zero cost. Powered by Pollinations.ai free API.
User says: "generate an image: a Shiba Inu on a beach drinking coconut water"
The agent performs the following workflow:
Extract the image description from user input.
Trigger keywords: "generate", "create an image", "draw", "make a picture"
ParameterDefaultDescriptionwidth1024Image width in pxheight768Image height in pxmodelsana``sana / flux / turbo / dreamshapernologotrueRemove watermarkseedrandomFixed integer for reproducible results
User can override:
text
"Generate: cyberpunk Shanghai night view, width 1920 height 1080, use flux model"
"Generate another, use seed=42 for consistent style"
Use exec to run curl against Pollinations.ai:
bash
PROMPT="<your text description>"
WIDTH=1024
HEIGHT=768
SEED="" # optional: "42"
curl -s --max-time 180 \
"https://image.pollinations.ai/prompt/$(python3 -c "import urllib.parse; print(urllib.parse.quote('$PROMPT'))")?width=$WIDTH&height=$HEIGHT&nologo=true${SEED:+&seed=$SEED}" \
-o /tmp/generated_image.jpg
Return the generated image file to the user.
ModelFeaturesSpeedsana (default)Latest efficient model, balanced style & detailFastfluxHigher quality, richer detailModerateturboFastest generation, good for previewsFastestdreamshaperArtistic / dream-like styleMedium
As of May 2026, Pollinations.ai default model has been switched to Sana. Older models like
stable-diffusionare deprecated. Sana offers better speed-quality balance.
text
User: "Generate: cyberpunk Shanghai night, neon lights, flying cars"
→ Agent calls API → returns image ✓
User: "Generate: an orange cat wearing an astronaut helmet, cartoon style, use flux model"
→ Agent calls API → returns image ✓
seed parameter (e.g., &seed=42) for reproducible results一个免费的AI图片生成工具。用户输入描述性文字,即可生成对应的图片。 无需任何API Key,零成本使用。基于 Pollinations.ai 免费API。
用户说:"画一张图:一只柴犬在沙滩上喝椰子水"
Agent 执行以下流程:
从用户输入中提取图片描述文本。
关键词触发:"画一张图"/"生成图片"/"画"/"create"/"generate"
参数默认值说明width1024图片宽度height768图片高度modelsana``sana / flux / turbo / dreamshaper无水印是默认去除水印seed随机固定值可复现相同结果
用户可指定参数:
text
"画一张图:赛博朋克上海夜景,宽1920高1080,使用flux模型"
"再生成一次,用seed=42保持风格一致"
使用 exec 运行 curl 命令调用 Pollinations.ai:
bash
PROMPT="<图片描述文字>"
WIDTH=1024
HEIGHT=768
SEED="" # 可选:"42"
curl -s --max-time 180 \
"https://image.pollinations.ai/prompt/$(python3 -c "import urllib.parse; print(urllib.parse.quote('$PROMPT'))")?width=$WIDTH&height=$HEIGHT&nologo=true${SEED:+&seed=$SEED}" \
-o /tmp/generated_image.jpg
将生成的图片文件返回给用户。
模型特点速度sana(默认)最新高效模型,通用风格,细节出色快flux更高质量,细节丰富稍慢turbo极速生成,适合快速预览最快dreamshaper偏向艺术/梦幻风格中等
2026年5月更新:Pollinations.ai 已将默认模型切换为 Sana,前代模型(如stable-diffusion)已下线。Sana在速度和画质间取得更好平衡。
text
用户:「画一张图:赛博朋克上海夜景,霓虹灯,飞行汽车」
→ Agent 调用API → 返回图片 ✓
用户:「画一张图:一只橘猫戴着宇航员头盔,卡通风格,使用flux模型」
→ Agent 调用API → 返回图片 ✓
seed 参数(如 &seed=42)获得可复现结果