Trackio
Trackio is a Python library for recording machine-learning experiments, with a dashboard that groups runs so you can compare metrics and configuration.
What Trackio is
Trackio is a Python library for recording machine-learning experiments, with a dashboard that groups runs so you can compare metrics and configuration. It runs its dashboard locally by default and can keep logs locally, in a Hugging Face Bucket, or on a self-hosted server. Its logging API follows the patterns of wandb.init, wandb.log, and wandb.finish, and is documented as a drop-in option for compatible workflows. You can log tables, Markdown reports, images, video, audio, 3D objects, HTML, and Matplotlib or Plotly figures. Optional packages add NVIDIA GPU and Apple Silicon system monitoring, with compatible hardware metrics logged automatically by default. Trackio documents integrations with Transformers and TRL, plus RapidFire AI for fine-tuning and RAG experiments. Alerts can be displayed, queried, stored, or sent through Slack and Discord webhooks. The dashboard can expose an HTTP API and MCP server; mutation tools require a write token. Trackio is free, but does not provide a first-class hyperparameter sweep API.
Good
- Free library and Hugging Face hosting.
- Dashboard groups runs for comparison.
- Logs support varied media and report types.
- Integrates with Transformers and TRL.
- Optional NVIDIA and Apple Silicon monitoring.
Watch out
- No first-class hyperparameter sweep API.
- Self-hosted ingestion and uploads require a secret write token.
Who it suits
Trackio suits Python users who want to record and compare autonomous ML experiments locally or on a self-hosted server. It may also fit teams using its documented integrations or alert options.
Verdict
Trackio brings experiment logging, run comparison, and alerting together in a free Python library. Users who need sweeps must manage configurations with Python loops, and self-hosting requires careful handling of its write token.
Plans and prices
From Trackio's own pricing page, checked 2 Oct 2026.
| Plan | Price | What it gives you |
|---|---|---|
| Free | Free | Trackio library and Hugging Face hosting are free; no plan limits stated |
The facts
Each one read from Trackio's own pages, with the date it was checked.
What it does
- Alerts
- Alerts can be printed, stored, queried, shown in the dashboard, and optionally sent to Slack or Discord webhooks.checked 2 Oct 2026
- API and MCP
- The dashboard can run as an HTTP API and MCP server, exposing read tools and mutation tools gated by a write token.checked 2 Oct 2026
- Experiment data
- Runs can log metrics and configuration, and the dashboard groups runs to compare experiments.checked 2 Oct 2026
- Integrations
- Trackio documents native integrations with Transformers and TRL, plus a RapidFire AI integration for fine-tuning and RAG experiments.checked 2 Oct 2026
- Intended users
- Trackio describes itself as designed for autonomous ML experiments and provides CLI commands and Python APIs for agents to log and query experiment data.checked 2 Oct 2026
- Logging
- Its API is compatible with wandb.init, wandb.log, and wandb.finish, and the docs show it can be used as a drop-in replacement.checked 2 Oct 2026
- Media
- The logging API supports tables, Markdown reports, images, video, audio, 3D objects, HTML, and Matplotlib or Plotly figures.checked 2 Oct 2026
- Notable limitation
- Trackio intentionally does not include a first-class hyperparameter sweep API; its docs describe sweeps as Python loops with one run per configuration.checked 2 Oct 2026
- Purpose
- Trackio is a lightweight Python library for tracking machine learning experiments, built on Hugging Face Buckets and Spaces.checked 2 Oct 2026
- Self-hosting security
- For a self-hosted server outside Hugging Face Spaces, metric ingestion and uploads require a write token, which the docs say to treat as a secret on untrusted networks.checked 2 Oct 2026
- Storage and dashboard
- Trackio runs a dashboard locally by default and can store logs locally, in a Hugging Face Bucket, or on a self-hosted server.checked 2 Oct 2026
- System monitoring
- Optional packages enable NVIDIA GPU monitoring and Apple Silicon system monitoring, with compatible hardware metrics logged automatically by default.checked 2 Oct 2026
Alternatives to Trackio
- 1
Paperspace Gradient
Free planOpen sourceWebAPIExtensionLinux8.3From $8 a month - 2
Weights & Biases
Free planAndroidiOSWebAPI8.3From $60 a month - 3
Comet
Free planWebAPILinuxmacOS7.9From $19 a month - 4
ClearML
Free planWebAPILinuxmacOS7.4From $15 a month - 5
MLRun
Free planWebAPILinuxmacOS6.6Free plan - 6
TensorBoard
Free planWebAPILinuxmacOS6.5Free plan