Model

Gemini Chrome (bridge)

Talk to Google Gemini 2.5 Pro through the OpenAI shape. Behind the model id gemini-chrome is a session-bridge that drives a real, authenticated Gemini browser tab on Meridian infrastructure — no Google API key required from you.

Pricing

Python

from openai import OpenAI

client = OpenAI(
    api_key="nim_live_***",
    base_url="https://meridian.getnimbus.net/api/v1",
)

resp = client.chat.completions.create(
    model="gemini-chrome",
    messages=[{"role": "user", "content": "Summarize the Linux kernel scheduler."}],
    # Bridge round-trip is 10-15s typical — raise client timeout
    timeout=60,
)
print(resp.choices[0].message.content)

cURL

curl https://meridian.getnimbus.net/api/v1/chat/completions \
  -H "Authorization: Bearer nim_live_***" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-chrome",
    "messages": [{"role": "user", "content": "ping"}]
  }'

When to pick it

• You want Gemini 2.5 Pro without managing your own Google API quota
• You want 1M-token context windows for long-doc reasoning
• You can tolerate 10–15s response times

When NOT to pick it

• Latency-sensitive workloads — use gemini-3-flash or gemini-3.1-pro instead (sub-2s).
• High-RPS bulk inference — bridge throughput is bounded by browser session count.
• Strict region pinning — bridge runs in Sweden Central only.