Google: Gemini 4 Argon
Google frontier model for coding, knowledge work, multimodal understanding, and cybersecurity defense. Announced September 30, 2026; access currently limited to trusted cyber defenders and testers, with broader API and Google AI Ultra availability planned. Announced introductory rate $2 input / $10 output per MTok, with cached input 95% cheaper ($0.10/MTok); $4/$20 after the introductory period (end date not announced). Output limit 1M tokens; the input context limit has not been published.
Sources
Same model, different sources. Our first-party rate is listed first, then router and aggregator rows stay separate so you can see the spread instead of a single blended number.
| Source | Provider | Input | Output | Cache read | Context |
|---|---|---|---|---|---|
| tokenstatcheapestgoogle/gemini-4-argon | $2 | $10 | $0.1 | N/A | |
| LiteLLMgoogle/gemini-4-argon | N/A | $2 | $10 | $0.1 | N/A |
| LiteLLMgemini-4-argon | N/A | $2 | $10 | $0.1 | N/A |
Capabilities
- Reasoning
- Tool calling
- Attachments
- In: text
- In: image
- In: video
- Out: text
Scores
Sparse coverage on purpose. Missing a score means we have not linked one yet, not that the model is unranked. Hover or tap the i next to any score for its source. Full boards on /benchmarks.
- DeepSWE v1.177.9i
- Harvey LAB (Vals)19.6i
- Terminal-Bench 4.057.4i
- OSWorld 2.0 (offline)69.2i
- Chartography (no tools)71.6i
- Vals Index68.9i
- AutomationBench (Private)51.3i
- Vals Finance Agent v265.4i
- FrontierSWE v255i
- Vibe Code Bench91.9i
- PostTrainBench v1.145.3i
- Terminal-Bench Science 0.1 (6x verifier timeout)57.6i
- LABBench 288.8i
- RiemannBench76i
- GraphWalks BFS F1 (up to 128K)99.7i
- GraphWalks BFS F1 (256K–1M)84.2i
- Agents' Last Exam (pass rate)39.5i
- LVBench (no tools)91.7i
- CWE-Bench v168i
Arena rows are community-vote Elo from Arena AI, Artificial Analysis indices arrive via OpenRouter, and launch-eval figures are collected from vendor announcements, analyst writeups, and community benchmarks. Full boards and credits on /benchmarks.
See what this model costs you
The app reads the sessions your tools already wrote and prices this model's tokens at the published API rate. Free on your computer.