DeepSeek V4 Pro

DeepSeek V4 Pro is the open-weight flagship in the V4 family, the successor to V3.2. It ships with a 1M-token context window, a native reasoning ('thinking') mode that can be toggled per request, function calling, and structured-output support. Designed for agentic coding, multi-step logical reasoning, mathematics, STEM, and competitive programming. Use it when you want frontier reasoning quality with open weights and an Anthropic-compatible API surface alongside the standard OpenAI shape.

Quick start

Use any OpenAI-compatible client. Just change base_url and your key.

from openai import OpenAI

client = OpenAI(
    api_key="tl-xxxxxxxxxxxxxxxxxxxxxxxx",
    base_url="https://api.thalam.ai/v1",
)

response = client.chat.completions.create(
    model="deepseek/deepseek-v4-pro",
    messages=[
        {"role": "user", "content": "Hello from Thalam!"}
    ],
)

print(response.choices[0].message.content)