$ diff k2-horizon gemini-3-8-flash
K2 Horizon vs Gemini 3.8 Flash
Values come from K2 Horizon and Gemini 3.8 Flash, 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.
IfmK2 Horizon
- context
- 524K
- weights
- open
- $/M in
- —
- $/M out
- —
- context
- 1.0M
- weights
- closed
- $/M in
- $0.75
- $/M out
- $3.75
What actually differs
- Context window
- Gemini 3.8 Flash takes 1.0M against 524K — 2× more room in a single request.
- Weights
- K2 Horizon publishes weights (Apache 2.0 (weights and code)); Gemini 3.8 Flash is API-only. That decides self-hosting, air-gapped deployment and fine-tuning before any capability question does.
- Recency
- K2 Horizon shipped 1 day after Gemini 3.8 Flash (2026-09-03 vs 2026-09-02).
Full spec
| Attribute | K2 Horizon | Gemini 3.8 Flash |
|---|---|---|
| Developer | Ifm | Google DeepMind |
| Released | 2026-09-03 | 2026-09-02 |
| Context window | 524,288 (flagship 375B-A23B, native from midtraining onward) | 1,048,576 tokens (65,536 max output) |
| Pricing | not recorded | $0.75/M input · $3.75/M output (introductory, through 2026-12-31) · $1.50/M · $7.50/M from 2027-01-01 |
| License | Apache 2.0 (weights and code) | proprietary |
| Availability | Hugging Face (weights, code, checkpoints, training data where redistributable) | Gemini API, Google AI Studio, Antigravity, Android Studio, Gemini Enterprise / Gemini Enterprise Agent Platform |