Llama 4 Maverick

Llama 4 Maverick is Meta's open-weight mixture-of-experts model in the Llama 4 family, with 17B active parameters across 128 experts and native multimodal (text and image) support. It's tuned for high-quality general assistant workloads, coding, and visual reasoning, served here in FP8 for efficient inference. A strong open-weight pick when you need multimodal capability plus Llama-ecosystem tooling.

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="meta-llama/llama-4-maverick-17b-128e-instruct-fp8",
    messages=[
        {"role": "user", "content": "Hello from Thalam!"}
    ],
)

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