$ diff embeddinggemma-2 k2-horizon
EmbeddingGemma 2 vs K2 Horizon
Values come from EmbeddingGemma 2 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 DeepMindEmbeddingGemma 2
- context
- 8K
- weights
- open
- $/M in
- —
- $/M out
- —
- context
- 524K
- weights
- open
- $/M in
- —
- $/M out
- —
What actually differs
- Context window
- K2 Horizon takes 524K against 8K — 65.5× more room in a single request.
- Weights
- Both publish weights — EmbeddingGemma 2 under Apache 2.0, K2 Horizon under Apache 2.0 (weights and code). Check the licences rather than assuming they permit the same commercial use.
- Recency
- EmbeddingGemma 2 shipped 33 days after K2 Horizon (2026-10-06 vs 2026-09-03).
Full spec
| Attribute | EmbeddingGemma 2 | K2 Horizon |
|---|---|---|
| Developer | Google DeepMind | Ifm |
| Released | 2026-10-06 | 2026-09-03 |
| Context window | 8K tokens | 524,288 (flagship 375B-A23B, native from midtraining onward) |
| Pricing | not recorded | not recorded |
| License | Apache 2.0 | Apache 2.0 (weights and code) |
| Availability | Weights on Hugging Face and Kaggle; local inference | Hugging Face (weights, code, checkpoints, training data where redistributable) |