NewsEditorialSource: Google AI for Developers
Google makes Gemini 4 Argon generally available in the Gemini API
Google made Gemini 4 Argon generally available in the Gemini API on October 13, 2026, according to its changelog. [1] Google lists gemini-4-argon as stable and describes its intended workloads as complex coding, agents, mathematics and science. [1][3] Its specifications list text, image, video, audio and PDF inputs, text output, a shared context limit of 1,048,576 tokens and a default synchronous output limit of 65,536 tokens. [3] Google lists prices through November 30, 2026, at $2 per million input tokens and $10 per million output tokens, including thinking tokens; on December 1, those rise to $4 and $20, respectively. [4] Search Engine Land corroborated the rollout announcement but supplied no independent access tests or benchmark results. [2]
Why it matters
Developers migrating applications should review request settings: Google says sampling parameters are ignored, caching is implicit only, and outputs above 65,536 tokens require background processing. [3][4] Function calling, structured outputs and code execution are supported, while computer use remains in preview. [3] Access needs clarification: Google’s API guide excludes the Free tier and Google AI Pro plan, while Search Engine Land reports Pro access in the Gemini app and Search, which may reflect different product entitlements. [2][4] Developers should verify API access for their accounts, test workload quality and budget for December’s higher token prices. [4]
Attribution
Sources
- 1Google AI for DevelopersRelease notes | Gemini APISource published (UTC)
- 2Search Engine LandGemini 4 Argon rolling out in Google AI Mode & GeminiSource published (UTC)
- 3Google AI for DevelopersGemini 4 Argon | Gemini API
- 4Google AI for DevelopersWhat's new in Gemini 4 Argon | Gemini API
Read the original source for full details and context.