Enable your LLM app to understand images with the phospho multimodal model. For optimal performance, this model is not censored or moderated. Ensuring this model is used in a safe way is your responsability.

​

Requirements

Create an account on phospho.ai and get your API key. You need to have setup a billing method. You can add a it in the Settings of your dashboard here.

​

Sending a request

To send a request, add:

• text: your text prompt. For instance: “What is this?”
• image_url: either a URL of the image or the base64 encoded image data. The inputs list must be of lenght 1.

Optionally, to better control the generation, you can specify the following optional parameters:

• max_new_tokens (int): default to 200. Max 250. The maximum number of tokens that can be generated in the response.
• temperature (float, between 0.1 and 1.0) Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
• repetition_penalty (float): Default to 1.15. This parameter helps in reducing the repetition of words in the generated content.
• top_p (float, between 0.0 and 1.0): Default to 1.0. This parameter controls the diversity of the response by limiting the possible next tokens to the top p percent most likely.

If you pass a URL, make sure it is a generally available image (for instance by passing the link in a Private Navigation window). To encode an image in base 64, you can use this website.

curl -X 'POST' \
  'https://api.phospho.ai/v2/predict' \
  -H 'accept: application/json' \
  -H 'Authorization: Bearer YOUR_PHOSPHO_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "inputs": [{"text": "What is this?", "image_url": "http://images.cocodataset.org/val2017/000000039769.jpg"}],
  "model": "phospho-multimodal"
}'

​

Pricing

The pricing is based on the number of images sent.

Note: You need to have a billing method setup to use the model. Acces your billing portal to add one.

Contact us for high volume pricing.