The llm-gemini 0.35 release supports Gemini 4 Argon, with Willison also documenting SVG examples, launch pricing and a planned price increase.
Simon Willison released llm-gemini 0.35 on October 13, adding support for Google’s new Gemini 4 Argon model to the plugin for his LLM command-line tool. The update gives developers using that plugin a way to access Argon alongside other models in Google’s Gemini family.
Willison said Gemini 4 Argon became available through the Gemini API on the same day, as well as to customers on Google’s $100-per-month subscription plan. His post announces the integration but does not provide installation instructions or explain the API access configuration required to use the model.
To illustrate the model’s output, Willison published three pelican SVG examples labeled low, medium and high. He calculated that the high example, featuring a pelican riding a bicycle, cost 17.166 cents to produce. These are individual creative demonstrations rather than a controlled or comprehensive model comparison.
Argon launched at $2 per million input tokens and $10 per million output tokens, according to Willison. He said those rates are scheduled to double on December 1. He also cautioned that headline token prices are difficult to compare directly because, in agentic systems, overall cost depends on how many tokens a model needs to complete a goal.
Willison additionally cited Artificial Analysis data showing a total cost of $2,407 to run its benchmark with Argon. He said he had checked the figures in a chart that he edited using ChatGPT so Argon’s information box would not overlap other numbers. The post does not describe the benchmark’s task composition or testing conditions, limiting what can be inferred from that figure alone.
For existing LLM users, the practical significance is that Gemini 4 Argon can now be incorporated into workflows already built around the llm-gemini plugin. Willison’s examples offer a concrete view of one creative use and its reported cost, but they do not establish how Argon will perform or what it will cost across unrelated workloads.
| Period | Input tokens | Output tokens |
|---|---|---|
| At launch | $2 per 1 million | $10 per 1 million |
| From December 1 | Twice the launch price | Twice the launch price |
Sources
Originally published by AI Feed: Simon Willison Adds Gemini 4 Argon to llm-gemini 0.35.
Top comments (0)