Multimodal Chat

Chat completion and image generation capabilities, including vision-enhanced chat and text-to-image generation on Subnet 19.

POSTv1/19/cortext/chat

Cortext Chat

Generate specialized responses using Cortext-based chat completion.

Required attributes

Name
model
Type
string
Description
The Cortext model to use (defaults to "cortext-ultra").
Name
messages
Type
array
Description
Array of message objects with role and content. Each message must include role (enum: "user", "assistant", "system") and content (string).

Optional attributes

Name
temperature
Type
number
Description
Controls randomness in the response (defaults to 0.0001).
Name
max_tokens
Type
integer
Description
Maximum number of tokens to generate (defaults to 4000).
Name
top_p
Type
number
Description
Controls diversity via nucleus sampling (defaults to 1).
Name
stream
Type
boolean
Description
Whether to stream the response (defaults to true).
Name
seed
Type
integer
Description
Random seed for reproducibility (defaults to 0).

This endpoint is authenticated. Get your API key

Request

POST

v1/19/cortext/chat

curl -X POST "https://api.bitmind.ai/oracle/v1/19/text/cortext/chat" \
  -H "Authorization: Bearer {token}" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cortext-ultra",
    "messages": [
      {
        "role": "user",
        "content": "What is machine learning?"
      }
    ],
    "temperature": 0.0001,
    "max_tokens": 100,
    "top_p": 1,
    "stream": false,
    "seed": 1
  }'

Response

[
  {
    "choices": [
      {
        "message": {
          "role": "user",
          "content": "Machine learning is a subset of artificial intelligence (AI) that involves the development of algorithms and statistical models that enable computers to perform tasks without explicit instructions. Instead of being programmed with specific rules, machine learning systems learn from data to identify patterns, make decisions, and improve their performance over time.\n\nKey aspects of machine learning include:\n\n1. **Data Input**: Machine learning models require large amounts of data to learn from. This data can be structured (like databases) or unstructured (like text and"
        },
        "finish_reason": "stop"
      }
    ],
    "system_fingerprint": null,
    "id": "chatcmpl-tEGXiayjTs_QuCyYdRhYyA",
    "created": 1739374983,
    "usage": {
      "prompt_tokens": 25,
      "completion_tokens": 101,
      "total_tokens": 126
    },
    "object": "chat.completion",
    "model": "gpt-4o"
  }
]

POSTv1/19/cortext/text-to-image

Cortext Text to Image

Generate images from text descriptions using Cortext models.

Required attributes

Name
prompt
Type
string
Description
The prompt used to create the image.
Name
model
Type
string
Description
The model used to generate the image (defaults to "cortext-image").
Name
style
Type
string
Description
Optional style parameter.
Name
size
Type
string
Description
The dimensions of the created image (enum: "256x256", "512x512", "1024x1024", defaults to "1024x1024").
Name
quality
Type
string
Description
The quality/resolution of the generated images (enum: "hd", "standard", defaults to "standard").

Optional attributes

Name
steps
Type
integer
Description
Number of steps in the image generation process (allowed values: 5-10, defaults to 5).
Name
cfg_scale
Type
integer
Description
Classifier-Free Guidance Scale (defaults to 8).
Name
seed
Type
integer
Description
Random seed for reproducibility (defaults to 0).

This endpoint is authenticated. Get your API key

Request

POST

v1/19/cortext/text-to-image

curl -X POST "https://api.bitmind.ai/oracle/v1/19/text/cortext/text-to-image" \
  -H "Authorization: Bearer {token}" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A beautiful sunset over mountains",
    "model": "cortext-image",
    "style": "vivid",
    "size": "1024x1024",
    "quality": "hd",
    "steps": 30,
    "cfg_scale": 8,
    "seed": 0
  }'

Response

[
  {
    "messages": "Give me an image of a cute, phantom, cockapoo. Very cute, not too fluffy",
    "model": "cortext-image",
    "style": "vivid",
    "size": "1024x1024",
    "quality": "hd",
    "completion": null,
    "steps": 30,
    "sampler": "",
    "cfg_scale": 8,
    "seed": 0,
    "provider": "OpenAI",
    "prompt": "A beautiful sunset over mountains",
    "image_url": "https://corcel.b-cdn.net/3294af65-bb87-4be3-9c46-5f50b39213ab.webp"
  }
]

POSTv1/19/vision/chat

Vision Chat

Generate responses with vision-enhanced capabilities for image-related discussions.

Required attributes

Name
messages
Type
array
Description
Array of message objects with role and content. Each message must include role (enum: "user", "assistant", "system") and content (string).

Optional attributes

Name
model
Type
string
Description
The vision model to use (defaults to "llama-3").
Name
temperature
Type
number
Description
Controls randomness in the response (defaults to 0.1).
Name
max_tokens
Type
integer
Description
Maximum number of tokens to generate (defaults to 100).
Name
top_p
Type
number
Description
Controls diversity via nucleus sampling (defaults to 1).
Name
stream
Type
boolean
Description
Whether to stream the response (defaults to false).
Name
logprobs
Type
boolean
Description
Whether to include log probabilities (defaults to false).

This endpoint is authenticated. Get your API key

Request

POST

v1/19/vision/chat

curl -X POST "https://api.bitmind.ai/oracle/v1/19/vision/chat" \
  -H "Authorization: Bearer {token}" \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "Hello, how are you?"
      }
    ],
    "model": "llama-3",
    "temperature": 0.1,
    "max_tokens": 100,
    "top_p": 1,
    "stream": false,
    "logprobs": false
  }'

Response

{
  "id": "chatcmpl-Q3g9MriBTyCAh9KrbFVN9g",
  "object": "chat.completion",
  "created": 1739377516,
  "model": "llama-3-1-70b",
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "Hello! I'm just a computer program, so I don't have feelings, but thank you for asking! How can I assist you today?"
      }
    }
  ],
  "usage": {
    "prompt_tokens": 19,
    "completion_tokens": 64,
    "total_tokens": 83
  },
  "system_fingerprint": null
}

POSTv1/19/vision/text-to-image

Vision Text to Image

Generate images using the vision model's text-to-image capabilities.

Required attributes

Name
text_prompts
Type
array
Description
Array of prompt objects, each containing text (string) and optional weight (number).

Optional attributes

Name
cfg_scale
Type
integer
Description
Classifier free guidance scale (defaults to 2).
Name
steps
Type
integer
Description
Number of diffusion steps (defaults to 8).
Name
engine
Type
string
Description
The engine to use (defaults to "proteus").
Name
height
Type
integer
Description
Output image height in pixels (defaults to 1024).
Name
width
Type
integer
Description
Output image width in pixels (defaults to 1024).

This endpoint is authenticated. Get your API key

Request

POST

v1/19/vision/text-to-image

curl -X POST "https://api.bitmind.ai/oracle/v1/19/vision/text-to-image" \
  -H "Authorization: Bearer {token}" \
  -H "Content-Type: application/json" \
  -d '{
    "text_prompts": [
      {
        "text": "4 hedgehogs, wearing tuxedos, riding on the back of a crocodile."
      }
    ],
    "cfg_scale": 2,
    "steps": 8,
    "engine": "proteus",
    "height": 1024,
    "width": 1024
  }'

Response

{
  "signed_urls": [
    "https://corcel.b-cdn.net/df42dc09-6de4-4c38-b716-5a0f064304c0.webp"
  ]
}

The multimodal chat model

The multimodal chat model contains the chat completion results.

Properties

Name
choices
Type
array
Description
Array containing the completion messages, including:
- message.role: Role of the message sender
- message.content: The message content
- finish_reason: Reason why the completion finished
Name
system_fingerprint
Type
string | null
Description
System fingerprint for the completion, if available.
Name
id
Type
string
Description
Unique identifier for the completion.
Name
created
Type
integer
Description
Unix timestamp of when the completion was created.
Name
usage
Type
object
Description
Token usage statistics for the completion, including prompt:
- prompt_tokens: Number of tokens in the prompt
- completion_tokens: Number of tokens in the completion
- total_tokens: Total number of tokens in the completion
Name
object
Type
string
Description
Object type, always "chat.completion".
Name
model
Type
string
Description
The model used for completion.