LLM Releases
← Catalog

Gemini 4 Hydrogen

Available
Google DeepMindProprietary

Google DeepMind's fast, efficient Gemini 4 model for everyday coding and agentic workflows, released Nov 12 2026 alongside the frontier Gemini 4 Carbon and Argon. Built to run alone or as the execution/worker model next to Carbon (Google Antigravity's new Orchestrator switches between them). Google reports 51 on the Artificial Analysis Intelligence Index at its default high thinking level (54 at xhigh) and 57% on Terminal-Bench 4.1 versus 18.2% for Gemini 3.8 Flash. Four thinking levels (low, medium, high, xhigh); native 1M-token multimodal input (text, images, audio, video, PDFs) with up to 1M output tokens. Flat pricing of $1.50 / $7.50 per 1M input/output tokens across the full context. Available in Google AI Studio, the Gemini API, Gemini for business, Antigravity, the Gemini app and Google Search. Proprietary; size not disclosed.

Specifications

License
Proprietary · Proprietary
Weights
Not released
Architecture
unknown
Parameters
Undisclosed
Context window
1.0M tokens
Max output
—
Knowledge cutoff
—
Price (in / out, $/M)
$1.5 / $7.5
Modalities
TextVisionAudioVideo

Benchmarks

No benchmark scores recorded yet. Spotted some? Submit a correction.

Vendor-reported figures are claims until independently verified. See methodology.