saymaker-haystack

Haystack components that generate images and video on SayMaker: one API key runs Veo 3.1, Kling 3.0, Seedance 2.0, MiniMax H3, Nano Banana 2, GPT Image 2.5, Seedream 5.0 and the rest of the shelf on your own credits.

pip install saymaker-haystack
export SAYMAKER_API_KEY=sk-...   # create one at https://saymaker.ai/settings/apikeys

from saymaker_haystack import SayMakerImageGenerator, SayMakerVideoGenerator

SayMakerImageGenerator(aspect_ratio="16:9").run(prompt="a lighthouse in fog, long exposure")
# {"images": ["https://cdn.saymaker.ai/..."]}

SayMakerVideoGenerator(duration=6).run(prompt="the lighthouse beam sweeps across the fog")
# {"videos": ["https://cdn.saymaker.ai/..."]}

In a pipeline

Write the prompt with an LLM, render it with SayMaker:

from haystack import Pipeline
from haystack.components.builders import PromptBuilder
from haystack.components.generators import OpenAIGenerator
from saymaker_haystack import SayMakerImageGenerator

p = Pipeline()
p.add_component("brief", PromptBuilder("Write one image prompt for a blog cover about {{topic}}. Name light, material, framing."))
p.add_component("llm", OpenAIGenerator())
p.add_component("render", SayMakerImageGenerator(aspect_ratio="16:9"))
p.connect("brief", "llm")
p.connect("llm.replies", "render.prompt")

(llm.replies is a list; add a small adapter component that takes replies[0] if your Haystack version does not convert it.)

Components

Both serialize with to_dict / from_dict; the key is a Haystack Secret read from SAYMAKER_API_KEY.

SayMakerVideoGenerator defaults to minimax-h3-fast, the one a free account can run. Pass model="veo-3-1", "kling-3-0" or "seedance-2" on a plan or credit pack (pricing). Runs on a free account wait in a queue before they start; the default timeouts cover the longest wait.

Built on the saymaker client. MIT licensed. Made by SayMaker, the AI video generator agent.