Models & pricing / gemini-4-hydrogen · Google
Google

gemini-4-hydrogen API

Gemini 4 Hydrogen is Google’s multimodal model for everyday coding and agent tasks. It accepts text, images, audio, video and documents and produces text. Compare current XiuAI access and prices, then check protocol-specific inputs and tools.[1][2]

gemini-4-hydrogen

Model overview

Published provider specifications. Available features depend on the current API route.

API routes
Gemini GenerateContent / OpenAI-compatible API / openai-response / openai-response-compact / Anthropic Messages / openai-alpha-search

Where Hydrogen fits

Consider Hydrogen when a coding question needs to be analyzed alongside screenshots, documents or other source material. Explain each input’s role and ask the output to identify supporting evidence. For audio or video analysis, check important passages against the original instead of relying only on a generated summary.[1]

Google positions Hydrogen for everyday coding and agent work. Compare the required input types, turnaround and current cost for your task. The vendor’s speed and efficiency positioning is not a fixed latency or completion-rate guarantee for an XiuAI route.[1][2]

Multimodal inputs and output length

Hydrogen produces text. Audio or video input for understanding does not imply audio or video generation, and image input does not imply image output. Choose a model for the output you need, then check how the selected protocol represents each input type.[1]

Google distinguishes default synchronous output from long decode. The context window, default output limit and long-output feature are separate limits. A documented long-decode ceiling does not establish that limit for an ordinary synchronous XiuAI request. Define sections and application-side checkpoints for longer tasks.[1]

Understanding multimodal costs

Text length does not represent the metered usage of every input. Images, audio, video and document processing can affect request usage. Estimate from the current prices and conditions on this page, then check actual charges in request records. Verify caching for the protocol and integration rather than assuming every vendor cache feature is enabled.[1][2]

API access

Using Hydrogen through XiuAI

Use the exact model ID gemini-4-hydrogen with an authorized XiuAI API key. Select a protocol from the catalog. Native Gemini contents/parts and compatibility-API messages use different structures, with separate handling for files and tools. Check a small sample before expanding the input.[1][2]

Vendor features such as search, code execution and computer use depend on particular tools and platform configuration. XiuAI model access does not automatically provide those environments. Your application must check tool support and handle execution, task state and output review.[1][2]

Send a request using your existing API key.

POST /v1/chat/completions
curl 'https://router-api.xiu.ai/v1/chat/completions' \
  -H "Authorization: Bearer $XIUAI_API_KEY" \
  -H 'Content-Type: application/json' \
  --data-raw '{
  "model": "gemini-4-hydrogen",
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ]
}'
Read the API documentation

Frequently asked questions

Does video understanding mean video generation?

No. Hydrogen’s documented output is text; video can be supplied for understanding. To generate video, choose a listed video-generation model and its asynchronous task protocol.[1][2]

Does the long-decode limit apply to every request?

No. Google documents default synchronous output and long decode separately, and availability depends on the API and integration. Do not copy a long-output ceiling into an ordinary request or treat the entire context window as generated text capacity.[1]

Can I send the same request through Gemini and compatibility APIs?

Convert the request structure and check capability differences. The same model ID does not imply identical message, image, file or tool fields. Confirm the available protocol in the catalog and construct its request format rather than changing only the URL.[1][2]

Related models

gemini-4-argongemini-4-carbongemini-3.8-flashgemini-3.1-flash-image

Sources and updates

Specifications reference provider documentation; prices and availability come from the current XiuAI catalog.

Content reviewed: 2026-11-16

[1] Google Gemini 4 Hydrogen documentation[2] Current XiuAI model and pricing data