# lean-agent-cruiser - leanmodels.ai

*Cruiserweight class, v1.0, DeepSeek V4-Flash base*

Million-token context - frontier long-context work on a single GPU

284B total, 13B active per token. DeepSeek's hybrid attention pairs Compressed Sparse Attention with Heavily Compressed Attention to hold a 1,048,576-token context, and Manifold-Constrained Hyper-Connections stabilise signal propagation across its 43 layers.

## Specifications

- **Total params:** 284B
- **Active per token:** 13B
- **Base model:** DeepSeek-V4-Flash
- **Architecture:** Hybrid CSA/HCA attention MoE
- **Experts:** 256 experts, 6 active + 1 shared
- **Target VRAM / RAM:** 24 GB / 64 GB
- **Tier:** Paid
- **License:** MIT

## Not yet verified

This model hasn't been benchmarked or verified on our hardware yet. Sizes and specs are targets, not measured results. No download or purchase until it clears the same verification the available models passed.

## Get the model

- **Download (target):** ~155.1 GB (single .lmpack, UD-Q4_K_XL)