$ diff deepseek-v4-1-flash k2-horizon
DeepSeek V4.1-Flash vs K2 Horizon
Values come from DeepSeek V4.1-Flash 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.
DeepSeekDeepSeek V4.1-Flash
- context
- 1M
- weights
- open
- $/M in
- $0.15
- $/M out
- $0.6
- context
- 524K
- weights
- open
- $/M in
- —
- $/M out
- —
What actually differs
- Context window
- DeepSeek V4.1-Flash takes 1M against 524K — modestly more room in a single request.
- Weights
- Both publish weights — DeepSeek V4.1-Flash under MIT (open-weight), K2 Horizon under Apache 2.0 (weights and code). Check the licences rather than assuming they permit the same commercial use.
- Recency
- DeepSeek V4.1-Flash shipped 7 days after K2 Horizon (2026-09-10 vs 2026-09-03).
Full spec
| Attribute | DeepSeek V4.1-Flash | K2 Horizon |
|---|---|---|
| Developer | DeepSeek | Ifm |
| Released | 2026-09-10 | 2026-09-03 |
| Context window | 1,000,000 | 524,288 (flagship 375B-A23B, native from midtraining onward) |
| Pricing | $0.15/M input (cache miss) · $0.003/M input (cache hit) · $0.60/M output, off-peak; peak rates double — see Pricing | not recorded |
| License | MIT (open-weight) | Apache 2.0 (weights and code) |
| Availability | DeepSeek API, Hugging Face (`deepseek-ai/DeepSeek-V4.1-Flash`) | Hugging Face (weights, code, checkpoints, training data where redistributable) |