---
id: cosmos-xenna
version: "0.5.7"
license: unclear
license_treatment: unclear
maintenance: active
---

# cosmos-xenna — A framework for building and running distributed, AI-powered data pipelines using Ray

License: unclear · Maintenance: active · Downloads: 132.7K/mo

## What it is and what it does

Cosmos-xenna is a distributed pipeline framework built on Ray that simplifies the development of AI inference workflows. You define a series of stages—each implementing setup() for one-time initialization and process_data() for batch processing—and connect them in a PipelineSpec. The framework handles resource allocation (CPUs, GPUs), autoscaling to balance throughput across stages, backpressure management to prevent memory overflow, and real-time monitoring. It supports three execution modes: streaming (recommended, all stages concurrent with dynamic worker balancing), batch (sequential, simpler but materializes intermediate data), and serving (online real-time input/output via queues). Most code is Python; autoscaling and artifact distribution logic is written in Rust for performance. The framework is designed for production workloads like video captioning, image processing, and other multi-model inference chains.

Use it for:

- Build multi-stage video or image processing pipelines that download data, run VLM inference, and upload results at scale
- Orchestrate complex AI workflows combining multiple models (e.g., detection → captioning → embedding) with automatic GPU/CPU allocation
- Process large datasets through streaming pipelines with automatic worker balancing to maximize throughput without manual tuning
- Deploy online serving scenarios where inference requests arrive in real time and results are pushed to output queues
- Manage resource-constrained inference by defining per-stage batch sizes and resource requirements to prevent OOM

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Cosmos-xenna is a Python framework for building and running distributed AI inference pipelines on Ray clusters, handling resource allocation, autoscaling, and data flow across multi-stage workflows.

Yes, with conditions. Cosmos-xenna is actively maintained and well-suited for production distributed AI pipelines on Ray. However, the package description explicitly states it is no longer under active development and recommends migration to Cosmos 3 for new projects. Install if you have an existing Cosmos-Xenna pipeline or need its specific feature set; otherwise, evaluate Cosmos 3 first. License terms are unclear and must be verified before use. No known security vulnerabilities as of August 2026.

## Install

pip install cosmos-xenna
uv add cosmos-xenna
poetry add cosmos-xenna

## Installing cosmos-xenna

Before you install: Medium install friction due to 9 runtime dependencies including Ray, a distributed computing framework, plus compiled wheels for x86_64 and aarch64. Active maintenance as of August 2026, though the package description notes that Cosmos-Xenna is no longer under active development and users are encouraged to migrate to Cosmos 3.

License in practice: License treatment is unclear—no SPDX identifier or raw license text is available in the package metadata. Verify the actual license before use in proprietary or restricted-license projects.

Quickstart:

pip install cosmos-xenna

import cosmos_xenna.pipelines.v1 as pipelines_v1

class MyStage(pipelines_v1.Stage):
    @property
    def stage_batch_size(self) -> int:
        return 10
    @property
    def required_resources(self) -> pipelines_v1.Resources:
        return pipelines_v1.Resources(gpus=0, cpus=1.0)
    def process_data(self, samples):
        return samples

pipeline_spec = pipelines_v1.PipelineSpec(
    input_data=samples,
    stages=[MyStage()]
)
pipelines_v1.run_pipeline(pipeline_spec)

Requires Python >= 3.12. Ray cluster must be running or accessible; cosmos-xenna orchestrates work across Ray actors.

Verify before relying:

- Actual license terms and restrictions (SPDX/raw license not provided in metadata)
- Whether Cosmos 3 migration path affects long-term support or breaking changes in cosmos-xenna
- Performance characteristics and throughput benchmarks for typical workloads
- Compatibility with Ray versions beyond what runtime dependency resolution enforces

## Package facts

- License: not declared (unclear)
- Python support: supports_current
- Install friction: medium
- Maintenance: active
- Downloads: 132.7K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

distributed AI pipeline orchestration, ray-based data processing framework, multi-stage inference pipeline, autoscaling distributed workflows, GPU-aware batch processing, streaming data pipeline framework, AI model inference at scale, ray-framework, ai-inference, pipeline-orchestration

[View on SkillFed](https://skillfed.io/packages/cosmos-xenna) · [View on PyPI](https://pypi.org/project/cosmos-xenna/)