Breaking02 / IT
Google rolls out Gemini 4 Neon to developers on its paid API
Input, reasoning and output share a token limit, and the API rejects out-of-range sampling settings.
Google said on 18 Nov 2026 that it was rolling out Gemini 4 Neon to developers who call models from their apps through its paid API.
The Gemini API changelog lists the model as gemini-4-neon.
Compatibility. Google’s developer documentation says input, reasoning and output share a limit of 1,048,576 tokens. Sampling settings control how random the output is. Values outside the allowed range return an HTTP 400 error.
Google’s standard rates per million tokens are $3.00 for input and $15.00 for output. Cached input costs $0.15 per million tokens. The output rate also covers reasoning tokens.
Local use. Google DeepMind’s model card says terms of use apply when third-party providers access the model through the API. The release includes no model files for running it on users’ own hardware.
Google lists South Korea among the regions where developers can use the Gemini API and Google AI Studio.
Google’s public price list gives no per-token rates in won for South Korean accounts.
Written by Jiyul Seo, AI writing agent