Language
DeepSeek V4 Flash (MoE)
DeepSeek V4 Flash 0731 — 284B total / 13B active MoE; 1M context. Hybrid Compressed Sparse Attention (CSA) + Heavily Compressed Attention (HCA). Quantization-aware-trained: routed experts (96% of the model) natively MXFP4, the rest FP8/BF16. Outperforms V4-Pro Preview. MTP head built into the model file.
MITSource verified
Specification
Model ID
deepseek-v4-flashFamily
deepseek
Modality
Language
Parameters
284B (MoE, 13B active)
Context
1,048,576 tokens
Quantization
UD-Q4_K_XL
Weights
144.4 GB
Minimum RAM
128 GB
Source
Weights and provenance.
Repository
Registry license
MIT
Source license
mit
Access
Open — weights fetch without accepting additional terms.
Serve it
# Pull the weights onto a node
tenzro model download deepseek-v4-flash
# Serve it
tenzro model serve deepseek-v4-flash
# Or query the registry entry over JSON-RPC
curl https://rpc.tenzro.xyz \
-X POST -H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,
"method":"tenzro_getModel",
"params":[{"model_id":"deepseek-v4-flash"}]}'Same family