Google Gemini 4 Hydrogen
Fast, low-cost execution tier of Gemini 4, built to work alone or under Gemini 4 Carbon/Argon as worker and verifier agents. Google reports 51 on the Artificial Analysis Intelligence Index at default high thinking (54 at xhigh) and 57% on Terminal-Bench 4.1 vs 18.2% for Gemini 3.8 Flash (vendor-reported).
Interactive Token Cost Estimator (Gemini 4 Hydrogen)
Architectural Strengths & Target Workloads
- Flat $1.50 / $7.50 per 1M tokens across a 1M-token window, $0.075 cache reads
- Batches file reads and searches in parallel; 12–23% fewer tokens than 3.8 Flash in partner tests (vendor-reported)
- 2.3x more resilient on Gray Swan indirect prompt injection than 3.8 Flash (vendor-reported)
Primary Engineering Workloads: Executor / Subagent Model · Repository Coding & Debugging · Multimodal Knowledge Work.
Frequently Asked Questions (Gemini 4 Hydrogen)
How much does Google Gemini 4 Hydrogen cost per 1M tokens?
Google Gemini 4 Hydrogen is priced at $1.50 per 1M uncached input tokens, $0.07 per 1M cached input tokens, and $7.50 per 1M output tokens (gemini-4-hydrogen on the Gemini API / AI Studio (free tier available), Gemini for business and Google Antigravity. Implicit caching only (4,096-token minimum prefix, no explicit cached_content). Batch/Flex $0.75 / $3.75; Priority $2.70 / $13.50.).
What is the context window of Google Gemini 4 Hydrogen?
It supports up to 1,048,576 tokens (flat pricing across the full window) (1,048,576 tokens) and a maximum generation output of 1,048,576 tokens.
Gemini 4 Hydrogen: Google's $1.50 Executor Model, Built to Work Under Carbon
Hydrogen is the fast, cheap tier of Gemini 4: 1M-token context, up to 1M output tokens, four thinking levels and flat $1.50 / $7.50 pricing. Google also launched Orchestrator in Antigravity, which switches between Carbon and Hydrogen inside one session.
Read Full Technical Dispatch (4 min read) →