GPT-6 Astra

GPT-6 Astra is OpenAI's newest frontier model and the most capable text model in our catalog. It carries a 1M-token context window, accepts text and images, and reports its reasoning tokens separately so you can see what the model spent before it answered. Reasoning effort is adjustable from none to high, and that setting matters here: reasoning tokens bill as output, so turning effort down on routine work is the difference between a cheap call and an expensive one. Function calling requires reasoning effort set to none. Reach for Astra on the hardest long-context work, and drop to the GPT-5.6 tiers once a workload proves it does not need the full model.

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="openai/gpt-6-astra",
    messages=[
        {"role": "user", "content": "Hello from Thalam!"}
    ],
)

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