Release Artifacts

This document provides a comprehensive inventory of all Dynamo release artifacts including container images, Python wheels, Helm charts, and Rust crates.

See also: Support Matrix for hardware and platform compatibility | Feature Matrix for backend feature support

Release history in this document begins at v0.6.0.

Current Release: Dynamo v1.0.2

• GitHub Release: v1.0.2
• Docs: v1.0.2
• NGC Collection: ai-dynamo

Experimental: v1.1.0-dev.3 (partial), v1.1.0-dev.2 (partial), and v1.1.0-dev.1 are available as experimental previews. Dev releases ship a subset of artifacts — see Pre-Release Artifacts for the exact images, wheels, and Helm charts published per version.

Container Images

* Multimodal inference on CUDA 13 images: works on AMD64 for all backends; works on ARM64 only for TensorRT-LLM (vllm-runtime:*-cuda13 and sglang-runtime:*-cuda13 do not support multimodality on ARM64).

Python Wheels

We recommend using the TensorRT-LLM NGC container instead of the ai-dynamo[trtllm] wheel. See the NGC container collection for supported images.

Helm Charts

Rust Crates

Quick Install Commands

Container Images (NGC)

$
# Runtime containers
$
docker pull nvcr.io/nvidia/ai-dynamo/vllm-runtime:1.0.2
$
docker pull nvcr.io/nvidia/ai-dynamo/sglang-runtime:1.0.2
$
docker pull nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:1.0.2
$
$
# CUDA 13 variants
$
docker pull nvcr.io/nvidia/ai-dynamo/vllm-runtime:1.0.2-cuda13
$
docker pull nvcr.io/nvidia/ai-dynamo/sglang-runtime:1.0.2-cuda13
$
$
# EFA variants (AWS, AMD64 only, experimental)
$
docker pull nvcr.io/nvidia/ai-dynamo/vllm-runtime:1.0.2-efa-amd64
$
docker pull nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:1.0.2-efa-amd64
$
$
# Infrastructure containers
$
docker pull nvcr.io/nvidia/ai-dynamo/dynamo-frontend:1.0.2
$
docker pull nvcr.io/nvidia/ai-dynamo/kubernetes-operator:1.0.2
$
docker pull nvcr.io/nvidia/ai-dynamo/snapshot-agent:1.0.2

Python Wheels (PyPI)

$
# Install Dynamo with a specific backend (Recommended)
$
uv pip install "ai-dynamo[vllm]==1.0.2"
$
uv pip install --prerelease=allow "ai-dynamo[sglang]==1.0.2"
$
# TensorRT-LLM requires the NVIDIA PyPI index and pip
$
pip install --pre --extra-index-url https://pypi.nvidia.com "ai-dynamo[trtllm]==1.0.2"
$
$
# Install Dynamo core only
$
uv pip install ai-dynamo==1.0.2
$
$
# Install standalone KVBM (Python 3.12 only)
$
uv pip install kvbm==1.0.2

Helm Charts (NGC)

$
helm install dynamo-platform oci://helm.ngc.nvidia.com/nvidia/ai-dynamo/charts/dynamo-platform --version 1.0.2
$
helm install snapshot oci://helm.ngc.nvidia.com/nvidia/ai-dynamo/charts/snapshot --version 1.0.2

Rust Crates (crates.io)

$
cargo add dynamo-runtime@1.0.2
$
cargo add dynamo-llm@1.0.2
$
cargo add dynamo-protocols@1.0.2
$
cargo add dynamo-parsers@1.0.2
$
cargo add dynamo-memory@1.0.2
$
cargo add dynamo-config@1.0.2
$
cargo add dynamo-tokens@1.0.2
$
cargo add dynamo-mocker@1.0.2
$
cargo add dynamo-kv-router@1.0.2

CUDA and Driver Requirements: For detailed CUDA toolkit versions and minimum driver requirements for each container image, see the Support Matrix.

Known Issues

For a complete list of known issues, refer to the release notes for each version:

• v1.0.2 Release Notes
• v1.0.1 Release Notes
• v1.0.0 Release Notes
• v0.9.0 Release Notes
• v0.8.1 Release Notes

Known Artifact Issues

Release History

• v1.1.0-dev.3 (experimental, partial): Preview release on release/1.1.0-dev.3. Ships only tensorrtllm-runtime:1.1.0-dev.3 (TRT-LLM v1.3.0rc11) plus ai-dynamo and ai-dynamo-runtime wheels. No vLLM/SGLang containers, no other component containers, no Helm charts published. Not recommended for production use.
• v1.1.0-dev.2 (experimental, partial): Preview release. Ships sglang-runtime:1.1.0-dev.2 (SGLang v0.5.9) and tensorrtllm-runtime:1.1.0-dev.2 (TRT-LLM v1.3.0rc9) plus ai-dynamo, ai-dynamo-runtime, and kvbm wheels. No vLLM container, no other component containers, no Helm charts published. Not recommended for production use.
• v1.1.0-dev.1 (experimental): Preview release. SGLang v0.5.9, TRT-LLM v1.3.0rc5.post1, vLLM v0.17.1, NIXL v0.10.1. Not recommended for production use.
• v1.0.2: Patch release. Same backend versions as v1.0.0: SGLang v0.5.9, TRT-LLM v1.3.0rc5.post1, vLLM v0.16.0, NIXL v0.10.1.
• v1.0.1: Patch release. Same backend versions as v1.0.0: SGLang v0.5.9, TRT-LLM v1.3.0rc5.post1, vLLM v0.16.0, NIXL v0.10.1.
• v1.0.0: First major release. SGLang v0.5.9, TRT-LLM v1.3.0rc5.post1 (CUDA 13.1), vLLM v0.16.0, NIXL v0.10.1. New snapshot-agent container and snapshot Helm chart (Preview). New EFA container variants for vLLM and TRT-LLM (Experimental, AMD64 only). New dynamo-mocker and dynamo-kv-router Rust crates. Deprecated dynamo-crds Helm chart (CRDs now managed by the Operator). v1alpha1 CRDs deprecated.
• v0.9.1: Updated TRT-LLM to v1.3.0rc3. All other backend versions unchanged from v0.9.0.
• v0.9.0.post1: Fixed dynamo-platform Helm chart operator image tag (Helm chart only, NGC)
• v0.9.0: Updated vLLM to v0.14.1, SGLang to v0.5.8, TRT-LLM to v1.3.0rc1, NIXL to v0.9.0. New dynamo-tokens Rust crate. Deprecated dynamo-graph Helm chart.
• v0.8.1.post1/.post2/.post3 Patches: Experimental patch releases updating TRT-LLM only (PyPI wheels and TRT-LLM container). No other artifacts changed.
• Standalone Frontend Container: dynamo-frontend added in v0.8.0
• EFA Runtimes: Experimental AWS EFA variants for vLLM and TRT-LLM (AMD64 only) in v1.0.0
• CUDA 13 Runtimes: Experimental CUDA 13 runtime for SGLang and vLLM in v0.8.0
• New Rust Crates: dynamo-memory and dynamo-config added in v0.8.0

GitHub Releases

Container Images

NGC Collection: ai-dynamo

To access a specific version, append ?version=TAG to the container URL: https://catalog.ngc.nvidia.com/orgs/nvidia/teams/ai-dynamo/containers/{container}?version={tag}

vllm-runtime

sglang-runtime

tensorrtllm-runtime

dynamo-frontend

kubernetes-operator

snapshot-agent

Python Wheels

PyPI: ai-dynamo | ai-dynamo-runtime | kvbm

To access a specific version: https://pypi.org/project/{package}/{version}/

ai-dynamo (wheel)

ai-dynamo-runtime (wheel)

kvbm (wheel)

Helm Charts

NGC Helm Registry: ai-dynamo

Direct download: https://helm.ngc.nvidia.com/nvidia/ai-dynamo/charts/{chart}-{version}.tgz

dynamo-crds (Helm chart) — Deprecated

dynamo-platform (Helm chart)

snapshot (Helm chart)

dynamo-graph (Helm chart) — Deprecated

Rust Crates

crates.io: dynamo-runtime | dynamo-llm | dynamo-protocols | dynamo-parsers | dynamo-memory | dynamo-config | dynamo-tokens

To access a specific version: https://crates.io/crates/{crate}/{version}

dynamo-runtime (crate)

dynamo-llm (crate)

dynamo-protocols (crate)

dynamo-parsers (crate)

dynamo-memory (crate)

dynamo-config (crate)

dynamo-tokens (crate)

dynamo-mocker (crate)

dynamo-kv-router (crate)

Pre-Release Artifacts

Pre-release Python wheels are published on the NVIDIA package index at pypi.nvidia.com, not on the public PyPI index. Like stable wheels, they are Linux (manylinux) builds for the Python versions in the Support Matrix; pip/uv on macOS or Windows will not find matching wheels. Install on a supported Linux host or inside a Linux container.

Install by adding that URL as an extra index and allowing pre-releases (PEP 440 dev versions):

$
# uv (recommended in other Dynamo docs)
$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo==1.1.0.dev2
$
$
# pip
$
pip install --pre --extra-index-url https://pypi.nvidia.com ai-dynamo==1.1.0.dev2

A GitHub or container tag v1.1.0-dev.N maps to a wheel version 1.1.0.devN (for example v1.1.0-dev.2 → ==1.1.0.dev2). Optional extras such as ai-dynamo[vllm] use the same flags; pin the version you want from the sections below.

v1.1.0-dev.3

• Branch: release/1.1.0-dev.3
• GitHub Tag: v1.1.0-dev.3
• Backends (branch ToT): SGLang v0.5.10.post1 | TensorRT-LLM v1.3.0rc11 | vLLM v0.19.0 | NIXL v0.10.1
• Coverage: Partial — only the TensorRT-LLM container and the ai-dynamo / ai-dynamo-runtime wheels are published. SGLang and vLLM containers, additional component containers (dynamo-frontend, kubernetes-operator, snapshot-agent), the kvbm wheel, and Helm charts are not published for this dev release.

Container Images

Python Wheels

Available from pypi.nvidia.com (pre-release index):

$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo==1.1.0.dev3
$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo-runtime==1.1.0.dev3

kvbm==1.1.0.dev3 is not yet published.

Helm Charts

Not published for this dev release. Use v1.1.0-dev.1 charts or the latest stable (v1.0.2) for platform install.

Rust Crates

Not shipped for pre-release versions.

v1.1.0-dev.2

• Branch: release/1.1.0-dev.2
• GitHub Tag: v1.1.0-dev.2
• Backends (branch ToT): SGLang v0.5.9 | TensorRT-LLM v1.3.0rc9 | vLLM v0.19.0 | NIXL v0.10.1
• Coverage: Partial — SGLang and TensorRT-LLM containers plus ai-dynamo, ai-dynamo-runtime, and kvbm wheels are published. vLLM container, additional component containers (dynamo-frontend, kubernetes-operator, snapshot-agent), and Helm charts are not published for this dev release.

Container Images

Python Wheels

Available from pypi.nvidia.com (pre-release index):

$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo==1.1.0.dev2
$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo-runtime==1.1.0.dev2
$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ kvbm==1.1.0.dev2

Helm Charts

Not published for this dev release. Use v1.1.0-dev.1 charts or the latest stable (v1.0.2) for platform install.

Rust Crates

Not shipped for pre-release versions.

v1.1.0-dev.1

• Branch: release/1.1.0-dev.1
• GitHub Tag: v1.1.0-dev.1
• Backends: SGLang v0.5.9 | TensorRT-LLM v1.3.0rc5.post1 | vLLM v0.17.1 | NIXL v0.10.1

Container Images

Python Wheels

Available from pypi.nvidia.com (pre-release index):

$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo==1.1.0.dev1
$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ ai-dynamo-runtime==1.1.0.dev1
$
uv pip install --pre --extra-index-url https://pypi.nvidia.com/ kvbm==1.1.0.dev1

Helm Charts

Rust Crates

Not shipped for pre-release versions.