$ diff gemini-4-argon hy4-preview
Gemini 4 Argon vs Hy4 preview
Values come from Gemini 4 Argon and Hy4 preview, where each is cited to its source. This page states no benchmark result and ranks neither model — it puts two published specifications next to each other. Where a lab has not published a figure, the row says so rather than guessing.
Google DeepMindGemini 4 Argon
- context
- 2M
- weights
- closed
- $/M in
- $2
- $/M out
- $10
- context
- 1M
- weights
- open
- $/M in
- $0.834
- $/M out
- $2.501
What actually differs
- Context window
- Gemini 4 Argon takes 2M against 1M — 2× more room in a single request.
- Input price
- Hy4 preview at $0.834/M against $2/M — 2.4× cheaper to feed. Standard rates: one of these labs also quotes a lower cached-input tier, which applies only when a prefix is reused — see the full spec below.
- Output price
- Hy4 preview at $2.501/M against $10/M — 4.0× cheaper to generate. Output dominates the bill on most agentic workloads, where the model writes far more than it reads.
- Weights
- Hy4 preview publishes weights (Apache-2.0); Gemini 4 Argon is API-only. That decides self-hosting, air-gapped deployment and fine-tuning before any capability question does.
- Recency
- Gemini 4 Argon shipped 33 days after Hy4 preview (2026-09-30 vs 2026-08-28).
Full spec
| Attribute | Gemini 4 Argon | Hy4 preview |
|---|---|---|
| Developer | Google DeepMind | Tencent |
| Released | 2026-09-30 | 2026-08-28 |
| Context window | 2M tokens | over 1,000,000 |
| Pricing | $2/M input · $10/M output · cached input $0.10/M (introductory; stated to double to $4/$20) | $0.834/M input · $2.501/M output |
| License | proprietary | Apache-2.0 |
| Availability | Fairwind Program (trusted cyber defenders) only; paid API and Google AI Ultra stated as next, no date | Hugging Face (BF16 + FP8), ModelScope, GitCode, CNB, Tencent Cloud TokenHub, OpenRouter, WorkBuddy, CodeBuddy, Yuanbao, ima |