One-line summary
On October 13, 2026, Google (Google DeepMind) released its flagship model Gemini 4 Argon to developers and enterprises worldwide, with introductory API pricing of $2 per million input tokens and $10 per million output tokens, and 95% off for cached input tokens, valid until November 30, 2026.
Event Overview
On October 13, 2026, Google announced the global availability of its flagship model Gemini 4 Argon in an official blog post signed by Koray Kavukcuoglu, SVP at Google DeepMind and Chief AI Architect at Google. The model targets complex, multi-step tasks in coding, finance, legal, cybersecurity and science, and ranks first on the Vals Index among other benchmarks, according to Google.
Key Information
| Item | Detail |
|---|---|
| Release date | October 13, 2026 |
| Publisher | Google (Google DeepMind) |
| Availability | AI Studio, Gemini platform API, Google Antigravity (AI Ultra subscribers), Gemini app and AI Mode in Search (AI Ultra/Pro subscribers) |
| Introductory API pricing | $2 per million input tokens, $10 per million output tokens |
| Caching pricing | Cached input tokens at 95% off the input price |
| Offer end | November 30, 2026 (then $4 input / $20 output per million tokens) |
| Benchmark results | #1 on Vals Index; 87% on Embodied Reasoning (ERQA) benchmark |
| Internal usage | Agentic code submissions up more than 35%, code rollbacks down by nearly a third |
Google disclosed real-world results from Argon in internal and early customer use: helping the Chrome team refactor ~6 million lines of legacy browser feature code within Chromium's ~40 million-line codebase; accelerating Gemma 4 E2B end-to-end inference throughput on an A100 from ~37,900 to over 108,000 tokens per second (2.85x); and leading the Embodied Reasoning benchmark with an 87% score, including seven consecutive real-world cornhole tosses on the Apptronik Apollo humanoid via a video-to-sim-to-real pipeline.