Google · Flagship
Gemini 4 Argon New
Google calls it a flagship reasoning model
- Provider
- Family
- Gemini
- Model type
- Flagship
- Official release date
- Oct 13, 2026
- Release type
- General availability
- Status
- Generally available
- Previous version
- Gemini 3.1 Pro
- Input
- text, image, video, audio, pdf
- Output
- text
- Context window
- 1,048,576 tokens
- Max output
- 65,536 tokens
- API model id
gemini-4-argon- API price (input)
- $2 per 1M tokens
- API price (output)
- $10 per 1M tokens
- Price notes
- Introductory price through 2026-11-30; $4 input / $20 output from 2026-12-01. No free API tier. Output includes thinking tokens.
- Consumer apps
- Gemini app and AI Mode for Google AI Pro/Ultra subscribers (per press coverage of Google announcement)
- Developer API
- Gemini API, Google AI Studio, Gemini Enterprise/Vertex
- Free access
- No free API tier
- Paid access
- —
- Availability
- —
- Licence
- proprietary
Main improvements
- Google calls it a flagship reasoning model
- Long decode of up to ~1M output tokens
- Aimed at software engineering, maths/science and autonomous agents
Best use cases
- Complex software engineering
- Scientific problem solving
- Multi-step agents
Benchmarks
No benchmark results recorded. We only list scores with a source, benchmark version and date.
More from Google
- Gemini 3.7 Flash and 3.5 Flash deprecation Deprecated · Oct 8, 2026
- Nano Banana 2.1 Generally available · Oct 6, 2026
- Gemini 3.8 Flash TTS Generally available · Sep 22, 2026
- Gemini 3.8 Flash Generally available · Sep 2, 2026
- Gemini Omni Flash (1.1) Generally available · Aug 27, 2026
- Gemini 3.5 Flash-Lite Generally available · Jul 21, 2026