Release Artifacts

Container images, Python wheels, Helm charts, Rust crates, and release history
View as Markdown

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 v1.1.1

Experimental: v1.2.0-deepseek-v4-dev.2 (DeepSeek-V4-Flash / V4-Pro on Blackwell, vLLM + SGLang containers only) is available as an experimental preview. Tagged Pre-Releases and experimental builds are listed under Pre-Release Artifacts.

Container Images

Image:TagDescriptionBackendCUDAArchNGCNotes
vllm-runtime:1.1.1Runtime container for vLLM backendvLLM v0.19.0v12.9AMD64/ARM64NGC: vllm-runtime 1.1.1
vllm-runtime:1.1.1-cuda13Runtime container for vLLM backend (CUDA 13)vLLM v0.19.0v13.0AMD64/ARM64NGC: vllm-runtime 1.1.1-cuda13
vllm-runtime:1.1.1-efa-amd64Runtime container for vLLM with AWS EFAvLLM v0.19.0v12.9AMD64NGC: vllm-runtime 1.1.1-efa-amd64Experimental
sglang-runtime:1.1.1Runtime container for SGLang backendSGLang v0.5.10.post1v12.9AMD64/ARM64NGC: sglang-runtime 1.1.1
sglang-runtime:1.1.1-cuda13Runtime container for SGLang backend (CUDA 13)SGLang v0.5.10.post1v13.0AMD64/ARM64NGC: sglang-runtime 1.1.1-cuda13
tensorrtllm-runtime:1.1.1Runtime container for TensorRT-LLM backendTRT-LLM v1.3.0rc11v13.1AMD64/ARM64NGC: tensorrtllm-runtime 1.1.1
tensorrtllm-runtime:1.1.1-efa-amd64Runtime container for TensorRT-LLM with AWS EFATRT-LLM v1.3.0rc11v13.1AMD64NGC: tensorrtllm-runtime 1.1.1-efa-amd64Experimental
dynamo-frontend:1.1.1API gateway with Endpoint Prediction Protocol (EPP)AMD64/ARM64NGC: dynamo-frontend 1.1.1
dynamo-planner:1.1.1Standalone Planner image used by Profiler jobs and Planner podsAMD64/ARM64NGC: dynamo-planner 1.1.1
kubernetes-operator:1.1.1Kubernetes operator for Dynamo deploymentsAMD64/ARM64NGC: kubernetes-operator 1.1.1
snapshot-agent:1.1.1Snapshot agent for fast GPU worker recovery via CRIUAMD64/ARM64NGC: snapshot-agent 1.1.1Preview

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.

PackageDescriptionPythonPlatformPyPI
ai-dynamo==1.1.1Main package with backend integrations (vLLM, SGLang, TRT-LLM)3.103.12Linux (glibc v2.28+)PyPI: ai-dynamo 1.1.1
ai-dynamo-runtime==1.1.1Core Python bindings for Dynamo runtime3.103.12Linux (glibc v2.28+)PyPI: ai-dynamo-runtime 1.1.1
kvbm==1.1.1KV Block Manager for disaggregated KV cache3.103.12Linux (glibc v2.28+)PyPI: kvbm 1.1.1

Helm Charts

ChartDescriptionNGC
dynamo-platform-1.1.1Platform services (etcd, NATS) and Dynamo Operator for Dynamo clusterNGC Helm: dynamo-platform-1.1.1
snapshot-1.1.1Snapshot DaemonSet for fast GPU worker recoveryNGC Helm: snapshot-1.1.1

The dynamo-crds Helm chart is deprecated as of v1.0.0; CRDs are now managed by the Dynamo Operator. The dynamo-graph Helm chart is deprecated as of v0.9.0.

Rust Crates

CrateDescriptionMSRV (Rust)crates.io
dynamo-runtime@1.1.1Core distributed runtime libraryv1.82crates.io: dynamo-runtime 1.1.1
dynamo-llm@1.1.1LLM inference enginev1.82crates.io: dynamo-llm 1.1.1
dynamo-protocols@1.1.1Async OpenAI-compatible API clientv1.82crates.io: dynamo-protocols 1.1.1
dynamo-async-openai@1.0.2Deprecated legacy OpenAI client; use dynamo-protocolsv1.82crates.io: dynamo-async-openai 1.0.2
dynamo-parsers@1.1.1Protocol parsers (SSE, JSON streaming)v1.82crates.io: dynamo-parsers 1.1.1
dynamo-memory@1.1.1Memory management utilitiesv1.82crates.io: dynamo-memory 1.1.1
dynamo-config@1.1.1Configuration managementv1.82crates.io: dynamo-config 1.1.1
dynamo-tokens@1.1.1Tokenizer bindings for LLM inferencev1.82crates.io: dynamo-tokens 1.1.1
dynamo-mocker@1.1.1Inference engine simulator for benchmarkingv1.82crates.io: dynamo-mocker 1.1.1
dynamo-kv-router@1.1.1KV-aware request routing libraryv1.82crates.io: dynamo-kv-router 1.1.1

Quick Install Commands

Container Images (NGC)

For detailed run instructions, see the backend-specific guides: vLLM | SGLang | TensorRT-LLM

$# Runtime containers
$docker pull nvcr.io/nvidia/ai-dynamo/vllm-runtime:1.1.1
$docker pull nvcr.io/nvidia/ai-dynamo/sglang-runtime:1.1.1
$docker pull nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:1.1.1
$
$# CUDA 13 variants
$docker pull nvcr.io/nvidia/ai-dynamo/vllm-runtime:1.1.1-cuda13
$docker pull nvcr.io/nvidia/ai-dynamo/sglang-runtime:1.1.1-cuda13
$
$# EFA variants (AWS, AMD64 only, experimental)
$docker pull nvcr.io/nvidia/ai-dynamo/vllm-runtime:1.1.1-efa-amd64
$docker pull nvcr.io/nvidia/ai-dynamo/tensorrtllm-runtime:1.1.1-efa-amd64
$
$# Infrastructure containers
$docker pull nvcr.io/nvidia/ai-dynamo/dynamo-frontend:1.1.1
$docker pull nvcr.io/nvidia/ai-dynamo/dynamo-planner:1.1.1
$docker pull nvcr.io/nvidia/ai-dynamo/kubernetes-operator:1.1.1
$docker pull nvcr.io/nvidia/ai-dynamo/snapshot-agent:1.1.1

Python Wheels (PyPI)

For detailed installation instructions, see the Local Quick Start in the README.

$# Install Dynamo with a specific backend (Recommended)
$uv pip install "ai-dynamo[vllm]==1.1.1"
$uv pip install --prerelease=allow "ai-dynamo[sglang]==1.1.1"
$# TensorRT-LLM requires the NVIDIA PyPI index and pip
$pip install --pre --extra-index-url https://pypi.nvidia.com "ai-dynamo[trtllm]==1.1.1"
$
$# Install Dynamo core only
$uv pip install ai-dynamo==1.1.1
$
$# Install standalone KVBM
$uv pip install kvbm==1.1.1

Helm Charts (NGC)

For Kubernetes deployment instructions, see the Kubernetes Installation Guide.

$helm install dynamo-platform oci://helm.ngc.nvidia.com/nvidia/ai-dynamo/charts/dynamo-platform --version 1.1.1
$helm install snapshot oci://helm.ngc.nvidia.com/nvidia/ai-dynamo/charts/snapshot --version 1.1.1

Rust Crates (crates.io)

For API documentation, see each crate on docs.rs. To build Dynamo from source, see Building from Source.

$cargo add dynamo-runtime@1.1.1
$cargo add dynamo-llm@1.1.1
$cargo add dynamo-protocols@1.1.1
$# Deprecated legacy crate name — pin only if a dependency requires it; new code should use dynamo-protocols:
$# cargo add dynamo-async-openai@1.0.2
$cargo add dynamo-parsers@1.1.1
$cargo add dynamo-memory@1.1.1
$cargo add dynamo-config@1.1.1
$cargo add dynamo-tokens@1.1.1
$cargo add dynamo-mocker@1.1.1
$cargo add dynamo-kv-router@1.1.1

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:

Known Artifact Issues

VersionArtifactIssueStatus
v0.9.0dynamo-platform-0.9.0Helm chart sets operator image to 0.7.1 instead of 0.9.0.Fixed in v0.9.0.post1
v0.8.1vllm-runtime:0.8.1-cuda13Container fails to launch.Known issue
v0.8.1sglang-runtime:0.8.1-cuda13, vllm-runtime:0.8.1-cuda13Multimodality not expected to work on ARM64. Works on AMD64.Known limitation
v0.8.0sglang-runtime:0.8.0-cuda13CuDNN installation issue caused PyTorch v2.9.1 compatibility problems with nn.Conv3d, resulting in performance degradation and excessive memory usage in multimodal workloads.Fixed in v0.8.1 (#5461)

Release Artifact History

Each bullet is a delta to what ships on NGC / Helm / PyPI / crates.io: net-new crates, removed Helm charts, or image lines that split or appear on the registry. See the inventory tables above for full matrices.

Stable releases first (newest first). Pre-Release Git Tags (v*-dev.*, experimental tracks) are summarized below; per-tag images and wheels are spelled out in Pre-Release Artifacts.

For backend version pins, see the version-pins table above and the GitHub Releases table below.

Stable Releases

  • v1.1.1: Patch release. Same backend versions as v1.1.0: SGLang v0.5.10.post1 (NIXL v1.0.1), TRT-LLM v1.3.0rc11 (NIXL v0.10.1), vLLM v0.19.0 (NIXL v0.10.1).
  • v1.1.0: Images: Split Planner into its own dynamo-planner image on NGC for Profiler jobs and Planner pods; worker and runtime images no longer bundle Planner (artifact boundary change, not a new engine capability). Crates: First 1.y.z publication on crates.io for dynamo-protocols (multi-protocol types; dynamo-async-openai remains deprecated with final release 1.0.2).
  • v1.0.2 / v1.0.1: No artifact additions or removals versus v1.0.0.
  • v1.0.0: Images: snapshot-agent, EFA variants for vLLM and TRT-LLM (AMD64 only). Crates: First publish of dynamo-mocker, dynamo-kv-router. Helm: Added snapshot (preview); dropped deprecated dynamo-crds from the publish stream (CRDs owned by the Operator).
  • v0.9.1: No artifact additions or removals versus v0.9.0.
  • v0.9.0: Crates: First publish of dynamo-tokens. Helm: Dropped deprecated dynamo-graph from the publish stream.
  • v0.8.0: Images: dynamo-frontend, CUDA 13 variants for vLLM and SGLang. Crates: First publish of dynamo-memory, dynamo-config.

Dynamo Nightlies

  • New as of v1.1.0*: ai-dynamo and ai-dynamo-runtime — nightly builds from main publish wheels tagged *.devYYYYMMDD. Install with pip or uv using --pre and the same NVIDIA extra-index pattern as Pre-Release Artifacts.

* *.devYYYYMMDD versioning for nightly main wheels began Apr 24, 2026.

Pre-Release and Experimental Git Tags

  • v1.2.0-deepseek-v4-dev.2: Images: vllm-runtime:*-deepseek-v4-cuda13-dev.2, sglang-runtime:*-deepseek-v4-cuda12-dev.2, sglang-runtime:*-deepseek-v4-cuda13-dev.2. Helm / PyPI: Not published for this tag (see Pre-Release Artifacts).
  • v1.1.0-dev.3: Images: tensorrtllm-runtime:1.1.0-dev.3. Wheels: ai-dynamo, ai-dynamo-runtime on pypi.nvidia.com (see below).
  • v1.1.0-dev.2: Images: sglang-runtime:1.1.0-dev.2, tensorrtllm-runtime:1.1.0-dev.2. Wheels: ai-dynamo, ai-dynamo-runtime on pypi.nvidia.com (see below).
  • v1.1.0-dev.1: Images: vLLM, SGLang, TRT-LLM runtime matrix (CUDA 12 / 13 and EFA variants as listed), dynamo-frontend, kubernetes-operator, snapshot-agent. Wheels: ai-dynamo, ai-dynamo-runtime on pypi.nvidia.com. Helm: dynamo-platform, snapshot at 1.1.0-dev.1 (see below).

Helm-Only Patches

  • v0.9.0.post1: Republished dynamo-platform Helm chart only (operator image tag correction).

Backend-Only Patch Trains

  • v0.8.1.post1 / .post2 / .post3: Republished TRT-LLM runtime image and PyPI wheels only.

crates.io Rust Packages

These crates use repository https://github.com/ai-dynamo/dynamo.git. The table lists each crate’s first non-placeholder publication on crates.io (excluding reservation uploads named 0.0.0-prerelease.0). Dates are from the crates.io registry index.

CrateFirst Published VersionDate (crates.io)
dynamo-runtime0.1.02025-03-18
dynamo-llm0.2.02025-05-01
dynamo-async-openai0.4.12025-08-27
dynamo-parsers0.5.02025-09-18
dynamo-memory0.8.02026-01-15
dynamo-config0.8.02026-01-15
dynamo-tokens0.9.02026-02-12
dynamo-mocker1.0.02026-03-13
dynamo-kv-router1.0.02026-03-13
dynamo-protocols1.1.02026-05-04

dynamo-async-openai is deprecated; 1.0.2 is its final crates.io release. Use dynamo-protocols for new dependencies (crate).

dynamo-tokenizers exists only in the Dynamo workspace and is not published on crates.io.

GitHub Releases

VersionRelease DateGitHubDocsNotes
v1.2.0-deepseek-v4-dev.2May 1, 2026TagExperimental (DeepSeek-V4-Flash / V4-Pro Blackwell preview; vLLM + SGLang containers only)
v1.1.1May 5, 2026ReleaseDocs
v1.1.0May 1, 2026ReleaseDocs
v1.1.0-dev.3Apr 18, 2026TagPre-Release (TRT-LLM Runtime Image + Wheels; see Pre-Release Artifacts)
v1.1.0-dev.2Apr 9, 2026TagPre-Release (SGLang + TRT-LLM Runtime Images + Wheels; see Pre-Release Artifacts)
v1.1.0-dev.1Mar 17, 2026TagExperimental
v1.0.2Apr 22, 2026ReleaseDocs
v1.0.1Mar 16, 2026ReleaseDocs
v1.0.0Mar 12, 2026ReleaseDocs
v0.9.1Mar 4, 2026ReleaseDocs
v0.9.0Feb 11, 2026ReleaseArchived docs unavailable
v0.8.1Jan 23, 2026ReleaseArchived docs unavailable
v0.8.0Jan 15, 2026ReleaseArchived docs unavailable
v0.7.1Dec 15, 2025ReleaseArchived docs unavailable
v0.7.0Nov 26, 2025ReleaseArchived docs unavailable
v0.6.1Nov 6, 2025Release
v0.6.0Oct 28, 2025Release

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

Image:TagvLLMArchCUDANotes
vllm-runtime:1.1.1v0.19.0AMD64/ARM64v12.9
vllm-runtime:1.1.1-cuda13v0.19.0AMD64/ARM64v13.0
vllm-runtime:1.1.1-efa-amd64v0.19.0AMD64v12.9Experimental
vllm-runtime:1.1.0v0.19.0AMD64/ARM64v12.9
vllm-runtime:1.1.0-cuda13v0.19.0AMD64/ARM64v13.0
vllm-runtime:1.1.0-efa-amd64v0.19.0AMD64v12.9Experimental
vllm-runtime:1.0.2v0.16.0AMD64/ARM64v12.9
vllm-runtime:1.0.2-cuda13v0.16.0AMD64/ARM64v13.0
vllm-runtime:1.0.2-efa-amd64v0.16.0AMD64v12.9Experimental
vllm-runtime:1.0.1v0.16.0AMD64/ARM64v12.9
vllm-runtime:1.0.1-cuda13v0.16.0AMD64/ARM64v13.0
vllm-runtime:1.0.1-efa-amd64v0.16.0AMD64v12.9Experimental
vllm-runtime:1.0.0v0.16.0AMD64/ARM64v12.9
vllm-runtime:1.0.0-cuda13v0.16.0AMD64/ARM64v13.0
vllm-runtime:1.0.0-efa-amd64v0.16.0AMD64v12.9Experimental
vllm-runtime:0.9.1v0.14.1AMD64/ARM64v12.9
vllm-runtime:0.9.1-cuda13v0.14.1AMD64/ARM64v13.0Experimental
vllm-runtime:0.9.0v0.14.1AMD64/ARM64v12.9
vllm-runtime:0.9.0-cuda13v0.14.1AMD64/ARM64v13.0Experimental
vllm-runtime:0.8.1v0.12.0AMD64/ARM64v12.9
vllm-runtime:0.8.0v0.12.0AMD64/ARM64v12.9
vllm-runtime:0.8.0-cuda13v0.12.0AMD64/ARM64v13.0Experimental
vllm-runtime:0.7.0.post2v0.11.2AMD64/ARM64v12.8Patch
vllm-runtime:0.7.1v0.11.0AMD64/ARM64v12.8
vllm-runtime:0.7.0.post1v0.11.0AMD64/ARM64v12.8Patch
vllm-runtime:0.7.0v0.11.0AMD64/ARM64v12.8
vllm-runtime:0.6.1.post1v0.11.0AMD64/ARM64v12.8Patch
vllm-runtime:0.6.1v0.11.0AMD64/ARM64v12.8
vllm-runtime:0.6.0v0.11.0AMD64v12.8

sglang-runtime

Image:TagSGLangArchCUDANotes
sglang-runtime:1.1.1v0.5.10.post1AMD64/ARM64v12.9
sglang-runtime:1.1.1-cuda13v0.5.10.post1AMD64/ARM64v13.0
sglang-runtime:1.1.0v0.5.10.post1AMD64/ARM64v12.9
sglang-runtime:1.1.0-cuda13v0.5.10.post1AMD64/ARM64v13.0
sglang-runtime:1.0.2v0.5.9AMD64/ARM64v12.9
sglang-runtime:1.0.2-cuda13v0.5.9AMD64/ARM64v13.0
sglang-runtime:1.0.1v0.5.9AMD64/ARM64v12.9
sglang-runtime:1.0.1-cuda13v0.5.9AMD64/ARM64v13.0
sglang-runtime:1.0.0v0.5.9AMD64/ARM64v12.9
sglang-runtime:1.0.0-cuda13v0.5.9AMD64/ARM64v13.0
sglang-runtime:0.9.1v0.5.8AMD64/ARM64v12.9
sglang-runtime:0.9.1-cuda13v0.5.8AMD64/ARM64v13.0Experimental
sglang-runtime:0.9.0v0.5.8AMD64/ARM64v12.9
sglang-runtime:0.9.0-cuda13v0.5.8AMD64/ARM64v13.0Experimental
sglang-runtime:0.8.1v0.5.6.post2AMD64/ARM64v12.9
sglang-runtime:0.8.1-cuda13v0.5.6.post2AMD64/ARM64v13.0Experimental
sglang-runtime:0.8.0v0.5.6.post2AMD64/ARM64v12.9
sglang-runtime:0.8.0-cuda13v0.5.6.post2AMD64/ARM64v13.0Experimental
sglang-runtime:0.7.1v0.5.4.post3AMD64/ARM64v12.9
sglang-runtime:0.7.0.post1v0.5.4.post3AMD64/ARM64v12.9Patch
sglang-runtime:0.7.0v0.5.4.post3AMD64/ARM64v12.9
sglang-runtime:0.6.1.post1v0.5.3.post2AMD64/ARM64v12.9Patch
sglang-runtime:0.6.1v0.5.3.post2AMD64/ARM64v12.9
sglang-runtime:0.6.0v0.5.3.post2AMD64v12.8

tensorrtllm-runtime

Image:TagTRT-LLMArchCUDANotes
tensorrtllm-runtime:1.1.1v1.3.0rc11AMD64/ARM64v13.1
tensorrtllm-runtime:1.1.1-efa-amd64v1.3.0rc11AMD64v13.1Experimental
tensorrtllm-runtime:1.1.0v1.3.0rc11AMD64/ARM64v13.1
tensorrtllm-runtime:1.1.0-efa-amd64v1.3.0rc11AMD64v13.1Experimental
tensorrtllm-runtime:1.0.2v1.3.0rc5.post1AMD64/ARM64v13.1
tensorrtllm-runtime:1.0.2-efa-amd64v1.3.0rc5.post1AMD64v13.1Experimental
tensorrtllm-runtime:1.0.1v1.3.0rc5.post1AMD64/ARM64v13.1
tensorrtllm-runtime:1.0.1-efa-amd64v1.3.0rc5.post1AMD64v13.1Experimental
tensorrtllm-runtime:1.0.0v1.3.0rc5.post1AMD64/ARM64v13.1
tensorrtllm-runtime:1.0.0-efa-amd64v1.3.0rc5.post1AMD64v13.1Experimental
tensorrtllm-runtime:0.9.1v1.3.0rc3AMD64/ARM64v13.0
tensorrtllm-runtime:0.9.0v1.3.0rc1AMD64/ARM64v13.0
tensorrtllm-runtime:0.8.1.post3v1.2.0rc6.post3AMD64/ARM64v13.0Patch
tensorrtllm-runtime:0.8.1.post1v1.2.0rc6.post2AMD64/ARM64v13.0Patch
tensorrtllm-runtime:0.8.1v1.2.0rc6.post1AMD64/ARM64v13.0
tensorrtllm-runtime:0.8.0v1.2.0rc6.post1AMD64/ARM64v13.0
tensorrtllm-runtime:0.7.0.post2v1.2.0rc2AMD64/ARM64v13.0Patch
tensorrtllm-runtime:0.7.1v1.2.0rc3AMD64/ARM64v13.0
tensorrtllm-runtime:0.7.0.post1v1.2.0rc3AMD64/ARM64v13.0Patch
tensorrtllm-runtime:0.7.0v1.2.0rc2AMD64/ARM64v13.0
tensorrtllm-runtime:0.6.1-cuda13v1.2.0rc1AMD64/ARM64v13.0Experimental
tensorrtllm-runtime:0.6.1.post1v1.1.0rc5AMD64/ARM64v12.9Patch
tensorrtllm-runtime:0.6.1v1.1.0rc5AMD64/ARM64v12.9
tensorrtllm-runtime:0.6.0v1.1.0rc5AMD64/ARM64v12.9

dynamo-frontend

Image:TagArchNotes
dynamo-frontend:1.1.1AMD64/ARM64
dynamo-frontend:1.1.0AMD64/ARM64
dynamo-frontend:1.0.2AMD64/ARM64
dynamo-frontend:1.0.1AMD64/ARM64
dynamo-frontend:1.0.0AMD64/ARM64
dynamo-frontend:0.9.1AMD64/ARM64
dynamo-frontend:0.9.0AMD64/ARM64
dynamo-frontend:0.8.1AMD64/ARM64
dynamo-frontend:0.8.0AMD64/ARM64Initial

kubernetes-operator

Image:TagArchNotes
kubernetes-operator:1.1.1AMD64/ARM64
kubernetes-operator:1.1.0AMD64/ARM64
kubernetes-operator:1.0.2AMD64/ARM64
kubernetes-operator:1.0.1AMD64/ARM64
kubernetes-operator:1.0.0AMD64/ARM64
kubernetes-operator:0.9.1AMD64/ARM64
kubernetes-operator:0.9.0AMD64/ARM64
kubernetes-operator:0.8.1AMD64/ARM64
kubernetes-operator:0.8.0AMD64/ARM64
kubernetes-operator:0.7.1AMD64/ARM64
kubernetes-operator:0.7.0.post1AMD64/ARM64Patch
kubernetes-operator:0.7.0AMD64/ARM64
kubernetes-operator:0.6.1AMD64/ARM64
kubernetes-operator:0.6.0AMD64/ARM64

dynamo-planner

Image:TagArchNotes
dynamo-planner:1.1.1AMD64/ARM64
dynamo-planner:1.1.0AMD64/ARM64New

snapshot-agent

Image:TagArchNotes
snapshot-agent:1.1.1AMD64/ARM64Preview
snapshot-agent:1.1.0AMD64/ARM64Preview
snapshot-agent:1.0.2AMD64/ARM64Preview
snapshot-agent:1.0.1AMD64/ARM64Preview
snapshot-agent:1.0.0AMD64/ARM64Preview

Python Wheels

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

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

ai-dynamo (wheel)

PackagePythonPlatformNotes
ai-dynamo==1.1.13.103.12Linux (glibc v2.28+)
ai-dynamo==1.1.03.103.12Linux (glibc v2.28+)
ai-dynamo==1.0.23.103.12Linux (glibc v2.28+)
ai-dynamo==1.0.13.103.12Linux (glibc v2.28+)
ai-dynamo==1.0.03.103.12Linux (glibc v2.28+)
ai-dynamo==0.9.13.103.12Linux (glibc v2.28+)
ai-dynamo==0.9.03.103.12Linux (glibc v2.28+)
ai-dynamo==0.8.1.post33.103.12Linux (glibc v2.28+)TRT-LLM v1.2.0rc6.post3
ai-dynamo==0.8.1.post13.103.12Linux (glibc v2.28+)TRT-LLM v1.2.0rc6.post2
ai-dynamo==0.8.13.103.12Linux (glibc v2.28+)
ai-dynamo==0.8.03.103.12Linux (glibc v2.28+)
ai-dynamo==0.7.13.103.12Linux (glibc v2.28+)
ai-dynamo==0.7.03.103.12Linux (glibc v2.28+)
ai-dynamo==0.6.13.103.12Linux (glibc v2.28+)
ai-dynamo==0.6.03.103.12Linux (glibc v2.28+)

ai-dynamo-runtime (wheel)

PackagePythonPlatformNotes
ai-dynamo-runtime==1.1.13.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==1.1.03.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==1.0.23.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==1.0.13.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==1.0.03.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.9.13.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.9.03.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.8.1.post33.103.12Linux (glibc v2.28+)TRT-LLM v1.2.0rc6.post3
ai-dynamo-runtime==0.8.1.post13.103.12Linux (glibc v2.28+)TRT-LLM v1.2.0rc6.post2
ai-dynamo-runtime==0.8.13.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.8.03.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.7.13.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.7.03.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.6.13.103.12Linux (glibc v2.28+)
ai-dynamo-runtime==0.6.03.103.12Linux (glibc v2.28+)

kvbm (wheel)

PackagePythonPlatformNotes
kvbm==1.1.13.103.12Linux (glibc v2.28+)
kvbm==1.1.03.103.12Linux (glibc v2.28+)
kvbm==1.0.23.103.12Linux (glibc v2.28+)
kvbm==1.0.13.103.12Linux (glibc v2.28+)
kvbm==1.0.03.103.12Linux (glibc v2.28+)
kvbm==0.9.13.103.12Linux (glibc v2.28+)
kvbm==0.9.03.103.12Linux (glibc v2.28+)
kvbm==0.8.13.103.12Linux (glibc v2.28+)
kvbm==0.8.03.103.12Linux (glibc v2.28+)
kvbm==0.7.13.103.12Linux (glibc v2.28+)
kvbm==0.7.03.103.12Linux (glibc v2.28+)Initial

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

The dynamo-crds Helm chart is deprecated as of v1.0.0. CRDs are now managed by the Dynamo Operator.

ChartNotes
dynamo-crds-0.9.1Last release
dynamo-crds-0.9.0
dynamo-crds-0.8.1
dynamo-crds-0.8.0
dynamo-crds-0.7.1
dynamo-crds-0.7.0
dynamo-crds-0.6.1
dynamo-crds-0.6.0

dynamo-platform (Helm chart)

ChartNotes
dynamo-platform-1.1.1
dynamo-platform-1.1.0
dynamo-platform-1.0.2
dynamo-platform-1.0.1
dynamo-platform-1.0.0
dynamo-platform-0.9.1
dynamo-platform-0.9.0-post1Helm fix: operator image tag
dynamo-platform-0.9.0
dynamo-platform-0.8.1
dynamo-platform-0.8.0
dynamo-platform-0.7.1
dynamo-platform-0.7.0
dynamo-platform-0.6.1
dynamo-platform-0.6.0

snapshot (Helm chart)

ChartNotes
snapshot-1.1.1Preview
snapshot-1.1.0Preview
snapshot-1.0.2Preview
snapshot-1.0.1Preview
snapshot-1.0.0Preview

dynamo-graph (Helm chart) — Deprecated

The dynamo-graph Helm chart is deprecated as of v0.9.0.
ChartNotes
dynamo-graph-0.8.1Last release
dynamo-graph-0.8.0
dynamo-graph-0.7.1
dynamo-graph-0.7.0
dynamo-graph-0.6.1
dynamo-graph-0.6.0

Rust Crates

crates.io: dynamo-runtime | dynamo-llm | dynamo-protocols | dynamo-async-openai (deprecated) | dynamo-parsers | dynamo-memory | dynamo-config | dynamo-tokens

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

dynamo-runtime (crate)

CrateMSRV (Rust)Notes
dynamo-runtime@1.1.1v1.82
dynamo-runtime@1.1.0v1.82
dynamo-runtime@1.0.2v1.82
dynamo-runtime@1.0.1v1.82
dynamo-runtime@1.0.0v1.82
dynamo-runtime@0.9.1v1.82
dynamo-runtime@0.9.0v1.82
dynamo-runtime@0.8.1v1.82
dynamo-runtime@0.8.0v1.82
dynamo-runtime@0.7.1v1.82
dynamo-runtime@0.7.0v1.82
dynamo-runtime@0.6.1v1.82
dynamo-runtime@0.6.0v1.82

dynamo-llm (crate)

CrateMSRV (Rust)Notes
dynamo-llm@1.1.1v1.82
dynamo-llm@1.1.0v1.82
dynamo-llm@1.0.2v1.82
dynamo-llm@1.0.1v1.82
dynamo-llm@1.0.0v1.82
dynamo-llm@0.9.1v1.82
dynamo-llm@0.9.0v1.82
dynamo-llm@0.8.1v1.82
dynamo-llm@0.8.0v1.82
dynamo-llm@0.7.1v1.82
dynamo-llm@0.7.0v1.82
dynamo-llm@0.6.1v1.82
dynamo-llm@0.6.0v1.82

dynamo-protocols (crate)

On crates.io, dynamo-protocols lists 1.1.0 as its first installable release (placeholder reservation 0.0.0-prerelease.0 omitted here like other 0.0.0-prerelease.* uploads). Earlier semver lines for the OpenAI-compatible client shipped under dynamo-async-openai — see #### dynamo-async-openai (crate) below.

CrateMSRV (Rust)Notes
dynamo-protocols@1.1.1v1.82
dynamo-protocols@1.1.0v1.82

dynamo-async-openai (crate)

Deprecated. Prefer dynamo-protocols. This crate remains published on crates.io for manifests pinned to the old package name.

CrateMSRV (Rust)Notes
dynamo-async-openai@1.0.2v1.82Final crates.io release
dynamo-async-openai@1.0.1v1.82
dynamo-async-openai@1.0.0v1.82
dynamo-async-openai@0.9.1v1.82
dynamo-async-openai@0.9.0v1.82
dynamo-async-openai@0.8.1v1.82
dynamo-async-openai@0.8.0v1.82
dynamo-async-openai@0.7.1v1.82
dynamo-async-openai@0.7.0v1.82
dynamo-async-openai@0.7.0-post1v1.82
dynamo-async-openai@0.6.1v1.82
dynamo-async-openai@0.6.0v1.82
dynamo-async-openai@0.5.1v1.82
dynamo-async-openai@0.5.0v1.82
dynamo-async-openai@0.4.1v1.82

dynamo-parsers (crate)

CrateMSRV (Rust)Notes
dynamo-parsers@1.1.1v1.82
dynamo-parsers@1.1.0v1.82
dynamo-parsers@1.0.2v1.82
dynamo-parsers@1.0.1v1.82
dynamo-parsers@1.0.0v1.82
dynamo-parsers@0.9.1v1.82
dynamo-parsers@0.9.0v1.82
dynamo-parsers@0.8.1v1.82
dynamo-parsers@0.8.0v1.82
dynamo-parsers@0.7.1v1.82
dynamo-parsers@0.7.0v1.82
dynamo-parsers@0.6.1v1.82
dynamo-parsers@0.6.0v1.82

dynamo-memory (crate)

CrateMSRV (Rust)Notes
dynamo-memory@1.1.1v1.82
dynamo-memory@1.1.0v1.82
dynamo-memory@1.0.2v1.82
dynamo-memory@1.0.1v1.82
dynamo-memory@1.0.0v1.82
dynamo-memory@0.9.1v1.82
dynamo-memory@0.9.0v1.82
dynamo-memory@0.8.1v1.82
dynamo-memory@0.8.0v1.82Initial

dynamo-config (crate)

CrateMSRV (Rust)Notes
dynamo-config@1.1.1v1.82
dynamo-config@1.1.0v1.82
dynamo-config@1.0.2v1.82
dynamo-config@1.0.1v1.82
dynamo-config@1.0.0v1.82
dynamo-config@0.9.1v1.82
dynamo-config@0.9.0v1.82
dynamo-config@0.8.1v1.82
dynamo-config@0.8.0v1.82Initial

dynamo-tokens (crate)

CrateMSRV (Rust)Notes
dynamo-tokens@1.1.1v1.82
dynamo-tokens@1.1.0v1.82
dynamo-tokens@1.0.2v1.82
dynamo-tokens@1.0.1v1.82
dynamo-tokens@1.0.0v1.82
dynamo-tokens@0.9.1v1.82
dynamo-tokens@0.9.0v1.82Initial

dynamo-mocker (crate)

CrateMSRV (Rust)Notes
dynamo-mocker@1.1.1v1.82
dynamo-mocker@1.1.0v1.82
dynamo-mocker@1.0.2v1.82
dynamo-mocker@1.0.1v1.82
dynamo-mocker@1.0.0v1.82Initial

dynamo-kv-router (crate)

CrateMSRV (Rust)Notes
dynamo-kv-router@1.1.1v1.82
dynamo-kv-router@1.1.0v1.82
dynamo-kv-router@1.0.2v1.82
dynamo-kv-router@1.0.1v1.82
dynamo-kv-router@1.0.0v1.82Initial

Pre-Release Artifacts

Pre-Release artifacts do not go through QA validation. Pre-release versions are experimental previews intended for early testing and feedback. They may contain bugs, breaking changes, or incomplete features. Use stable releases for production workloads.

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.

v1.2.0-deepseek-v4-dev.2

  • Branch: release/1.2.0-deepseek-v4-dev.2
  • GitHub Tag: v1.2.0-deepseek-v4-dev.2
  • Backends: vLLM v0.20.0 (native DeepSeek-V4 support) | SGLang upstream lmsysorg/sglang:deepseek-v4-blackwell preview | NIXL v0.10.1
  • Coverage: DeepSeek-V4-Flash and V4-Pro only. vLLM and SGLang containers are published for Blackwell. TensorRT-LLM container, other component containers, Helm charts, and wheels are not published for this tag. Snapshot dev build for early-access V4 model support; not QA-gated.

Container Images

Image:TagBackendCUDAArch
vllm-runtime:1.2.0-deepseek-v4-cuda13-dev.2vLLM v0.20.0v13.0AMD64/ARM64
sglang-runtime:1.2.0-deepseek-v4-cuda12-dev.2SGLang upstream DSv4 previewv12.9AMD64
sglang-runtime:1.2.0-deepseek-v4-cuda13-dev.2SGLang upstream DSv4 previewv13.0ARM64

Python Wheels

Not published for this dev release. Use the v1.1.0 wheels or v1.1.0-dev.3 from pypi.nvidia.com.

Helm Charts

Not published for this dev release. Use v1.1.0 charts for platform install.

Rust Crates

Not shipped for pre-release versions.

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: TensorRT-LLM runtime container plus ai-dynamo and ai-dynamo-runtime wheels on pypi.nvidia.com. SGLang and vLLM containers, component containers (dynamo-frontend, dynamo-planner, kubernetes-operator, snapshot-agent), kvbm wheel, and Helm charts are not published for this tag.

Container Images

Image:TagBackendCUDAArch
tensorrtllm-runtime:1.1.0-dev.3TRT-LLM v1.3.0rc11v13.1AMD64/ARM64

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 the latest stable (v1.1.0) 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: SGLang and TensorRT-LLM runtime containers plus ai-dynamo and ai-dynamo-runtime wheels on pypi.nvidia.com. vLLM runtime container, component containers (dynamo-frontend, dynamo-planner, kubernetes-operator, snapshot-agent), kvbm wheel, and Helm charts are not published for this tag.

Container Images

Image:TagBackendCUDAArch
sglang-runtime:1.1.0-dev.2SGLang v0.5.9v12.9AMD64/ARM64
tensorrtllm-runtime:1.1.0-dev.2TRT-LLM v1.3.0rc9v13.1AMD64/ARM64

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

Helm Charts

Not published for this dev release. Use the latest stable (v1.1.0) for platform install.

Rust Crates

Not shipped for pre-release versions.

v1.1.0-dev.1

Container Images

Image:TagBackendCUDAArch
vllm-runtime:1.1.0-dev.1vLLM v0.17.1v12.9AMD64/ARM64
vllm-runtime:1.1.0-dev.1-cuda13vLLM v0.17.1v13.0AMD64/ARM64
vllm-runtime:1.1.0-dev.1-efa-amd64vLLM v0.17.1v12.9AMD64
sglang-runtime:1.1.0-dev.1SGLang v0.5.9v12.9AMD64/ARM64
sglang-runtime:1.1.0-dev.1-cuda13SGLang v0.5.9v13.0AMD64/ARM64
tensorrtllm-runtime:1.1.0-dev.1TRT-LLM v1.3.0rc5.post1v13.1AMD64/ARM64
tensorrtllm-runtime:1.1.0-dev.1-efa-amd64TRT-LLM v1.3.0rc5.post1v13.1AMD64
dynamo-frontend:1.1.0-dev.1AMD64/ARM64
kubernetes-operator:1.1.0-dev.1AMD64/ARM64
snapshot-agent:1.1.0-dev.1AMD64/ARM64

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

Helm Charts

ChartNGC
dynamo-platform-1.1.0-dev.1NGC Helm: dynamo-platform 1.1.0-dev.1
snapshot-1.1.0-dev.1NGC Helm: snapshot 1.1.0-dev.1

Rust Crates

Not shipped for pre-release versions.