$ diff gemini-4-argon k2-horizon
Gemini 4 Argon vs K2 Horizon
Values come from Gemini 4 Argon and K2 Horizon, 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
- 524K
- weights
- open
- $/M in
- —
- $/M out
- —
What actually differs
- Context window
- Gemini 4 Argon takes 2M against 524K — 3.8× more room in a single request.
- Weights
- K2 Horizon publishes weights (Apache 2.0 (weights and code)); 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 27 days after K2 Horizon (2026-09-30 vs 2026-09-03).
Full spec
| Attribute | Gemini 4 Argon | K2 Horizon |
|---|---|---|
| Developer | Google DeepMind | Ifm |
| Released | 2026-09-30 | 2026-09-03 |
| Context window | 2M tokens | 524,288 (flagship 375B-A23B, native from midtraining onward) |
| Pricing | $2/M input · $10/M output · cached input $0.10/M (introductory; stated to double to $4/$20) | not recorded |
| License | proprietary | Apache 2.0 (weights and code) |
| Availability | Fairwind Program (trusted cyber defenders) only; paid API and Google AI Ultra stated as next, no date | Hugging Face (weights, code, checkpoints, training data where redistributable) |