Gemini 4 Carbon and Hydrogen are live on Runware

Runware added Google's Gemini 4 Carbon and Gemini 4 Hydrogen to its OpenAI-compatible endpoint. The model IDs are google:gemini@4-carbon and google:gemini@4-hydrogen.

Carbon has a 1M-token context and up to 65,536 output tokens. It takes text, image, video, audio and PDF input and has configurable thinking. It costs $2 per million input tokens and $10 per million output tokens through November 30. From December 1 that becomes $4 input and $20 output.

Hydrogen also has a 1M-token context, with four thinking levels from low to xhigh, at $1.50 input and $7.50 output per million tokens. Google reports 57% on Terminal-Bench 4.1 for Hydrogen, up from 18.2% on Gemini 3.8 Flash.

Google pairs the two models. Carbon takes architecture and hard debugging, and Hydrogen runs the fast execution steps.