Google's Gemini 4 Argon now available on AI Gateway

Google's Gemini 4 Argon is now accessible via AI Gateway, designed specifically for coding, agentic workflows, and complex reasoning tasks. The model supports a 1M token context window and accepts diverse inputs including text, images, PDFs, and video, while enabling tool calling for extended operations.

Capabilities and Use Cases

Gemini 4 Argon excels at building features, fixing bugs, and researching unfamiliar codebases by carrying out multi-step tasks. It is also capable of interpreting dense diagrams, spreadsheets, and layout-heavy documents, making it suitable for financial analysis across multiple tabs. Beyond coding, the model handles complex problems in finance, health, law, math, and science.

Integration and Configuration

Developers can access the model using the identifier google/gemini-4-argon through the AI SDK, Chat Completions, Responses, and Anthropic Messages APIs. It also integrates with coding agents like Claude Code, Codex, and Cursor via the latest Vercel CLI. Instead of thinking budgets, the model utilizes thinking levels (low, medium, and high), which can be configured via reasoning in the AI SDK or reasoning_effort in Chat Completions.

Platform Features

AI Gateway offers a unified API for model calls, usage tracking, and performance optimizations like retries and failover. The platform reflects provider pricing with no markup and charges no platform fee on inference, including for Bring Your Own Key (BYOK) requests. Additional features include custom reporting, Zero Data Retention support, and budget controls for API keys.