Trust · Updated June 2026

Editorial standards for OpenSourcesAI.

OpenSourcesAI exists to help builders choose open-source AI tools, open-weight models, local LLM runtimes, RAG infrastructure, coding agents, and deployment stacks without confusing marketing claims for production guidance.

What we cover

We focus on tools and models that matter to AI builders: local model runners, model-serving runtimes, vector databases, RAG frameworks, coding assistants, agent tools, AI infrastructure, document workflows, data pipelines, and practical open-weight models.

We may also cover hosted frontier models for context, but the core directory is focused on open-source, open-weight, local, self-hosted, and developer-controlled workflows.

How we classify openness

  • Open source: source code is available under a recognized open-source license.
  • Open weight: model weights are available, but the training data, full source, or usage rights may still be restricted.
  • Open core: a usable open component exists, with paid or hosted features around it.
  • Source available: code or weights can be inspected, but terms may not meet open-source standards.
  • Commercial: proprietary hosted or paid software that can still be relevant to AI workflows.

Preferred sources

For factual model and tool claims, we prefer official project documentation, vendor docs, GitHub repositories, Hugging Face model cards, release notes, technical reports, license files, and primary provider announcements. Community posts and third-party reviews can help with discovery, but they should not replace primary sources for licenses, pricing, context windows, hardware requirements, or production suitability.

Verification dates

Some pages include updated or verified-through notes. These labels mean the page was reviewed against available public information at that point in time. They do not guarantee that pricing, licensing, model limits, context windows, runtime support, or hosted-provider terms have not changed since then.

Affiliate and sponsor policy

OpenSourcesAI may earn a commission from some partner links. Affiliate relationships do not guarantee positive coverage, ranking, inclusion, or recommendation. Sponsored placements should be clearly labeled, and editorial pages should still describe tradeoffs, limitations, and alternatives.

A commercial tool can be listed when it solves a real AI-builder workflow, but commercial status should be visible so readers can compare it against open-source or self-hosted options.

Corrections and updates

AI tools and models change quickly. If a model card, license, runtime, integration, or product page changes, the related OpenSourcesAI page should be updated rather than left as a stale claim. Readers and vendors can submit corrections through the community forum or sponsorship/contact paths.

Logo and trademark note

Tool, model, and integration logos or favicons are used only to help readers recognize projects and products in directory cards. Trademarks and logos belong to their respective owners. Inclusion does not imply endorsement by the trademark owner.