Updated

Image and video input on an abliterated LLM API

Send images as base64 in any of the three formats, and video as base64 in Chat Completions only. Kimi K3.1 Pro takes video, GLM-5.3 is text only, and the other models take images.

Chat Completions
https://api.silkcompute.com
reply = client.chat.completions.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    messages=[{"role": "user", "content": [
        {"type": "text", "text": "What does this error dialog say?"},
        {"type": "image_url", "image_url": {"url": f"data:image/png;base64,{png}"}},
    ]}],
)
message = client.messages.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    max_tokens=2048,
    messages=[{"role": "user", "content": [
        {"type": "image", "source": {"type": "base64", "media_type": "image/png", "data": png}},
        {"type": "text", "text": "What does this error dialog say?"},
    ]}],
)
response = client.responses.create(
    model="deepseek-v4.1-flash-1017-uncensored",
    store=False,
    input=[{"role": "user", "content": [
        {"type": "input_text", "text": "What does this error dialog say?"},
        {"type": "input_image", "image_url": f"data:image/png;base64,{png}"},
    ]}],
)
reply = client.chat.completions.create(
    model="kimi-k3.1-pro-uncensored",
    messages=[{"role": "user", "content": [
        {"type": "text", "text": "List the steps shown in this screen recording."},
        {"type": "video_url", "video_url": {"url": f"data:video/mp4;base64,{clip}"}},
    ]}],
)

How it works

ImagesBase64 in any formatChat Completions, Messages and Responses all take images.
VideoChat Completions onlySend a base64 video_url part to Kimi K3.1 Pro.
CheckModels can misreadCheck what a model reads off an image or clip.

Rules

CaseWhat happens
Video on MessagesReturns 400
Video on ResponsesDropped, still billed
Video on a new account403 video_not_available until it has some history
Request sizeUp to 64 MiB
BillingImages and video count as input tokens