Skip to content

DeepSeek

deepseek-ai/DeepSeek-V3-0324

DeepSeek V3 0324 MoE serving with pipeline-parallel deployment guidance.

Start this recipeView on Hugging Face
Model metadata
CategoryMoEParams37B/671BContext163,840BackendsCUDA, METALTasksChat
Recipe authorship · Recorded configuration verification

Recipe authorship: Curated

Curated and generated describe how this recipe was written, not whether it ran successfully.

Recorded configuration verification

No current matching configuration verification record.

Records cover the tested default configuration on the named hardware. Matching release minor, variant, and rendering hash are required; changing options does not verify the new configuration.

Choose your actual device. Your browser cannot detect GPU memory.

Command builder

Hardware
Variant

mlx-community/DeepSeek-V3-0324-4bit

Mode: Server
Advanced
Mode
Context

Context length is measured in tokens. Use the slider or enter a number.

Slots
Features
This feature is not available in the selected mode.
Memory estimation: Why unavailable?

This feature is not available in the selected mode.

Estimated memory use

approx

Approximate memory allocation in GB. These estimates are not measured peak usage.

  • Weights: ≈ 351.5 GB
  • KV: ≈ 13.3 GB
  • Drafter: ≈ 0.0 GB
  • Runtime allowance: ≈ 30.7 GB

Estimated to exceed the ≈ 384.0 GB available memory by ≈ 11.6 GB.

Adjust hardware or advanced settings

Notes

  • This tier is estimated from nearby measurements.

1. Install

Install in your terminal. Uncached model IDs download automatically on first use.

MLXCEL_INSTALL_DIR="${MLXCEL_INSTALL_DIR:-$HOME/.local/share/mlxcel}" \
  && mkdir -p "$MLXCEL_INSTALL_DIR" \
  && curl -fL -o "$MLXCEL_INSTALL_DIR/mlxcel.zip" https://github.com/lablup/mlxcel/releases/download/v0.8.1/mlxcel-macos-aarch64.zip \
  && unzip -o "$MLXCEL_INSTALL_DIR/mlxcel.zip" -d "$MLXCEL_INSTALL_DIR/mlxcel-macos-aarch64" \
  && export PATH="$MLXCEL_INSTALL_DIR/mlxcel-macos-aarch64:$PATH" \
  && mlxcel --version
Installation and download help

Run the install command in your terminal. The selected executable should then be available on PATH. If it is not found, repeat the displayed PATH export in the same terminal.

Model IDs download automatically when first used. The optional download command prepares the checkpoint ahead of time. Uncached models require internet access and enough disk space; gated models also require access approval and credentials.

2. Run

Run this command in the same terminal after preparation. For server modes, keep it running and wait for the listening/ready message. If memory is insufficient, select a smaller checkpoint or reduce context/concurrency in advanced settings.

Copy all includes the selected installation, checkpoint download, and displayed run commands, in that order. Run them in the same shell; multi-node commands belong on their respective machines.

mlxcel serve

This installation provides the full CLI. Use mlxcel serve; select a source build to use the standalone server.

mlxcel serve -m mlx-community/DeepSeek-V3-0324-4bit --alias deepseek-v3-0324 \
  --ctx-size 8192 --n-parallel 4
Command flag help
-m
Selects the model or checkpoint repository.
--alias
Sets the served model name used by API clients.
--ctx-size
Sets the maximum context window for requests.
--n-parallel
Sets the number of parallel server slots.

3. Check the first result

For a server, use a second terminal after it is ready. A text chat request returns JSON with choices[0].message.content; specialized endpoints return their task output. Replace sample file paths with your own inputs where shown. Connection refused: check the server address/port and terminal logs. Model access/download errors: check the model ID, permissions, internet connection, and free disk space.

curl http://127.0.0.1:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"deepseek-v3-0324","messages":[{"role":"user","content":"Hello"}]}'

Performance evidence

No benchmark is recorded for this selected checkpoint and hardware combination. Commands and memory estimates use the catalog's model information.

Memory fit is a model-based estimate, separate from measured throughput and successful inference verification.

No measurement is available for this hardware yet. See the family page for related profiles. Family

Prerequisites

  • Install the mlxcel build for your selected hardware and check the selected checkpoint's memory estimate.
  • Allow internet access and enough disk space to download an uncached checkpoint; obtain Hugging Face access approval when the checkpoint requires it.

Recipe options

  • • The fit estimate uses all 671B parameters because every expert weight must be resident even though only a subset is active per token.
  • • Pipeline memory estimates are approximate for each machine. The coordinator assigns layers at runtime.

Family

DeepSeek v3 / R1

MoE

Sibling recipes

pending

Guide

DeepSeek V3 0324

DeepSeek V3 is a large routed MoE. This recipe estimates memory use on one machine and offers pipeline parallelism to divide model layers across machines.

Plan the deployment

  1. Select the memory tier for each machine.
  2. Inspect the fit meter. A red single-machine result can still include a PP alternative.
  3. Apply the PP alternative or open the PP tab, then choose two to eight machines.
  4. Run the coordinator command on one machine and the peer command on every other machine.

The generated commands use mDNS to discover peers. All machines must share the same network and use compatible mlxcel builds.

Capacity notes

  • The memory estimate includes all 671B parameters, including inactive experts.
  • The per-machine estimate divides weights across the selected machine count while keeping KV cache and headroom local.
  • The coordinator assigns layers at runtime. Use the site's per-machine memory estimate to plan the configuration.
  • For static peer lists or manual layer partitions, use the pipeline-parallelism manual.