Chat | OpenDunes Docs

client.chat.completions.create() is the core method — it maps directly to POST /v1/chat/completions. Every model in the catalog is available through it.

Basic completion#

Python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://opendunes.com/api/v1",
    api_key=os.environ["OPENDUNES_API_KEY"],
)

resp = client.chat.completions.create(
    model="anthropic/claude-opus-4.8",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the capital of Algeria?"},
    ],
    max_tokens=256,
    temperature=0.7,
)
print(resp.choices[0].message.content)

The response object is a standard ChatCompletion. Check resp.usage for token counts — input and output tokens are both billed at the model's DA rate.

Streaming#

Set stream=True to receive delta chunks over SSE as they are generated. The stream closes with data: [DONE].

Python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://opendunes.com/api/v1",
    api_key=os.environ["OPENDUNES_API_KEY"],
)

stream = client.chat.completions.create(
    model="google/gemini-2.5-pro",
    messages=[{"role": "user", "content": "Explain recursion in plain Arabic."}],
    stream=True,
)

for chunk in stream:
    delta = chunk.choices[0].delta.content
    if delta:
        print(delta, end="", flush=True)
print()

The final chunk in the stream carries usage (prompt tokens, completion tokens). Your remaining DA balance is in the X-Balance-Available response header after each request.

Vision / image inputs#

Pass an array of content parts to include an image alongside text. The model must support the vision capability — filter by ?capability=vision in the model catalog.

Python

resp = client.chat.completions.create(
    model="anthropic/claude-opus-4.8",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Describe what is in this image."},
                {
                    "type": "image_url",
                    "image_url": {"url": "https://example.com/image.jpg"},
                },
            ],
        }
    ],
)
print(resp.choices[0].message.content)

Parameters#

Parameter	Type	Description
`model`	string	Required. Model slug, e.g. `openai/gpt-5.4`.
`messages`	array	Required. Array of `{role, content}` objects.
`stream`	bool	Stream tokens as SSE. Default `false`.
`max_tokens`	int	Cap on output tokens.
`temperature`	float	0–2. Higher = more varied output.
`top_p`	float	0–1. Nucleus sampling threshold.

Note. Tool/function calling (tools, tool_choice) and structured outputs (response_format) are supported and forwarded on every request — see the Tool Calling page.

Basic completion#

Python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://opendunes.com/api/v1",
    api_key=os.environ["OPENDUNES_API_KEY"],
)

resp = client.chat.completions.create(
    model="anthropic/claude-opus-4.8",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is the capital of Algeria?"},
    ],
    max_tokens=256,
    temperature=0.7,
)
print(resp.choices[0].message.content)

The response object is a standard ChatCompletion. Check resp.usage for token counts — input and output tokens are both billed at the model's DA rate.

Streaming#

Set stream=True to receive delta chunks over SSE as they are generated. The stream closes with data: [DONE].

Python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://opendunes.com/api/v1",
    api_key=os.environ["OPENDUNES_API_KEY"],
)

stream = client.chat.completions.create(
    model="google/gemini-2.5-pro",
    messages=[{"role": "user", "content": "Explain recursion in plain Arabic."}],
    stream=True,
)

for chunk in stream:
    delta = chunk.choices[0].delta.content
    if delta:
        print(delta, end="", flush=True)
print()

The final chunk in the stream carries usage (prompt tokens, completion tokens). Your remaining DA balance is in the X-Balance-Available response header after each request.

Vision / image inputs#

Pass an array of content parts to include an image alongside text. The model must support the vision capability — filter by ?capability=vision in the model catalog.

Python

resp = client.chat.completions.create(
    model="anthropic/claude-opus-4.8",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Describe what is in this image."},
                {
                    "type": "image_url",
                    "image_url": {"url": "https://example.com/image.jpg"},
                },
            ],
        }
    ],
)
print(resp.choices[0].message.content)

Parameters#

Parameter	Type	Description
`model`	string	Required. Model slug, e.g. `openai/gpt-5.4`.
`messages`	array	Required. Array of `{role, content}` objects.
`stream`	bool	Stream tokens as SSE. Default `false`.
`max_tokens`	int	Cap on output tokens.
`temperature`	float	0–2. Higher = more varied output.
`top_p`	float	0–1. Nucleus sampling threshold.

Note. Tool/function calling (tools, tool_choice) and structured outputs (response_format) are supported and forwarded on every request — see the Tool Calling page.