Gemini 4 Hydrogen API on BazaarLink: Price, Context and How to Call It

Gemini 4 Hydrogen is now available on BazaarLink. Google says it is built for rapid execution and suits agentic workflows. On BazaarLink it costs $1.50 per 1M input tokens and $7.50 per 1M output tokens, with a 1M-token context.

It uses the same interface as every other chat model, so an existing integration does not change: set model to the value below and try it.

Model and price

BazaarLink price (per 1M tokens, USD, read from the model catalog on 2026-11-13):

There was no limited-time discount at the time of writing. The model page always shows the current price.

What Google says

These are the claims in Google's announcement and DeepMind model card, published on 2026-11-12. BazaarLink has not tested them independently, and real speed and token use depend on your requests:

• Positioning: Google says Carbon and Argon excel at deep, long-horizon reasoning on the hardest problems, while Hydrogen is built for rapid execution, and calls it a "daily driver for agentic workflows".
• Google says that for teams migrating from Gemini 3.8 Flash, Hydrogen is a major upgrade: it finishes tasks 2.5 times faster at its default high thinking setting.
• Google says its Artificial Analysis Intelligence Index score at the default high thinking setting is 51 (Argon: 50), and 54 at xhigh.
• Google says its Terminal-Bench 4.1 score is 57%.
• Google says partners saw Hydrogen run 2x to 2.8x faster than Gemini 3.8 Flash in early testing, with 12% to 23% fewer tokens.

Hydrogen, Argon or Carbon

• For fast, high-volume agent and coding workflows, try Gemini 4 Hydrogen first.
• For deeper reasoning, Google says Gemini 4 Argon and Gemini 4 Carbon excel at deep, long-horizon reasoning; run the same prompts side by side.
• Current prices for all three are in the model catalog. They share one /api/v1 endpoint, so switching only means changing the model value.

Moving from Gemini 3.8 Flash

1. Change model in your request to gemini-4-hydrogen. The base URL and API key stay the same.
2. Reasoning effort can be low, medium, high or xhigh (default medium; reasoning is always on for this model).
3. Test output quality, speed and token use on your own data before switching everything.
4. If you send images, files, audio or video, check that your input types are in the table above.

How to call it

curl https://api.bazaarlink.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $BAZAARLINK_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-4-hydrogen",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

If anything behaves differently from this page, send us the request ID and we will look into it.

TWD billing · Taiwan invoices · leading AI models · OpenAI-compatible API