Simon Willison reported the release of llm-gemini 0.35, which introduces support for the new Gemini 4 Argon model. The model became available via the Gemini API and for customers using the $100 per month subscription plan. According to the report, Argon launched with pricing at $2 per million input tokens. Testing of the model produced SVG illustrations of pelicans across low, medium, and high settings, with the high-detail illustration costing 17.166 cents to generate.
API Access and Pricing Context
The update brings integration for Google’s newly released Argon model into the existing llm-gemini ecosystem. Gemini models have previously established a distinct reputation for generating Scalable Vector Graphics, with dedicated staff investment in that specific output format. The pricing structure set at $2 per million input tokens provides a baseline for developers evaluating API costs for the new architecture.
Open Questions and Future Details
The announcement did not specify the full capabilities of Gemini 4 Argon beyond API availability and the noted input token pricing. Furthermore, the publication did not detail output token pricing, context window limits, or specific performance benchmarks compared to earlier Gemini iterations. Observers will need to watch for future updates or direct provider documentation to understand the broader technical specifications of the Argon release.
Tooling updates like llm-gemini 0.35 allow developers to begin testing the new model directly in their existing command-line workflows.
Lead image: Bibek ghosh / Pexels
