#5757th of 57 in Gateways and observability

deepchecks

Python checks for validating tabular, NLP and vision data and models

Stars
4.1k
License
custom license
Last commit
Nov 2025
Last release
Dec 2024
Language
Python
Stars, last 29 weeks+81
25 Apr 2026: 3,988 stars2 May 2026: 3,992 stars9 May 2026: 3,995 stars16 May 2026: 3,998 stars23 May 2026: 4,001 stars30 May 2026: 4,003 stars6 Jun 2026: 4,007 stars13 Jun 2026: 4,011 stars20 Jun 2026: 4,014 stars27 Jun 2026: 4,015 stars4 Jul 2026: 4,021 stars11 Jul 2026: 4,026 stars18 Jul 2026: 4,030 stars25 Jul 2026: 4,030 stars1 Aug 2026: 4,033 stars8 Aug 2026: 4,033 stars15 Aug 2026: 4,040 stars22 Aug 2026: 4,041 stars29 Aug 2026: 4,043 stars5 Sept 2026: 4,046 stars12 Sept 2026: 4,049 stars19 Sept 2026: 4,050 stars26 Sept 2026: 4,052 stars3 Oct 2026: 4,057 stars10 Oct 2026: 4,058 stars17 Oct 2026: 4,062 stars24 Oct 2026: 4,065 stars31 Oct 2026: 4,067 stars7 Nov 2026: 4,068 stars14 Nov 2026: 4,069 stars

3,988 on 25 Apr 20264,069 on 14 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

Deepchecks is a Python package of built-in Checks and Suites that test data integrity, drift, leakage and model performance for tabular, NLP and computer vision projects. Results can be saved as HTML, shown in Jupyter, read as JSON or Python values, and gated with pass/fail conditions in CI. A separate repo, deepchecks/monitoring, reuses the same checks for production models.

Who it is for: ML engineers who need data and model validation tests in research and CI

Strengths

  • Built-in checks for drift, leakage, conflicting labels and weak segments across three data types
  • Checks take configurable parameters and conditions, and custom checks can be written
  • Results export to HTML, Jupyter, JSON or Python values for CI gating
  • Installs with pip or conda, with optional nlp and vision extras

Weaknesses

  • AGPL 3.0 licensing may block use in some commercial products
  • Managed CI and testing management is a closed preview, not open source
  • Self-hosted open-source monitoring supports only one model per deployment
  • Monitoring repo has a commercial-licensed ee directory; last release was Dec 2024

What it needs

  • Needs Docker (self-hosted monitoring only)
  • Models: Tabular, NLP, Computer vision

What changed

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Gateways and observability

See all 57
Also in Gateways and observability
RankProjectScore
1OmniRouteLLM gateway routing one OpenAI-compatible endpoint across 372 providers88.5k starsMIT96 out of 100
2headroomCompresses tool outputs, logs and history before they reach the LLM79.9k starsApache-2.092 out of 100
3new-apiGateway that unifies LLM providers behind OpenAI, Anthropic and Gemini APIs52.1k starsAGPL-3.0known advisory in latest version91 out of 100
#4MLflowTracing, evaluation and prompt registry for LLM apps, plus ML tracking Live demo ↗ (opens in a new tab)28.9k starsApache-2.091 out of 100
#5LiteLLMOpenAI-format gateway and Python SDK for calling 100+ LLM providers63.5k starscustom license88 out of 100