Open LLMs vs The Stack v2

Maintains a continuously updated, manually curated registry of open-source large language models with commercial-use licensing. The repository implements a structured catalog approach where each model entry includes metadata (model name, organization, parameter count, license type, release date, and commercial eligibility) organized in markdown tables and JSON structures, enabling developers to filter and discover models based on licensing constraints, model size, and use-case suitability without legal ambiguity.

Unique: Focuses specifically on commercial-use licensing eligibility rather than general model benchmarking or capability comparison — filters out models with restrictive licenses (e.g., research-only, non-commercial clauses) upfront, reducing legal risk for production deployments

vs alternatives: More legally-focused than Hugging Face Model Hub (which lists all models regardless of commercial restrictions) and more current than static LLM comparison papers, providing a practical filtering layer for compliance-conscious teams

Aggregates heterogeneous model metadata from multiple sources (model cards, GitHub repositories, research papers, official announcements) and normalizes it into a consistent schema with fields for model name, organization, parameter count, license, release date, and commercial-use status. The implementation uses markdown tables as the primary data structure with optional JSON exports, enabling both human-readable browsing and programmatic access through simple parsing.

Unique: Uses a deliberately simple, human-readable markdown-first schema rather than complex database structures, making the registry accessible to non-technical stakeholders while remaining machine-parseable for automation

vs alternatives: Simpler and more accessible than database-backed model registries (e.g., MLflow Model Registry) but less queryable; trades flexibility for transparency and ease of contribution

Implements a filtering mechanism that categorizes models by their license type and commercial-use permissions, distinguishing between fully commercial-eligible models (Apache 2.0, MIT, OpenRAIL-M) and restricted models (research-only, non-commercial clauses, or ambiguous licensing). The filtering is applied at the curation stage where models are manually reviewed against licensing criteria before inclusion in the registry.

Unique: Explicitly prioritizes commercial-use licensing as the primary filtering criterion rather than model performance or capability, addressing a specific pain point for enterprises that need legal certainty before deployment

vs alternatives: More legally-focused than general model discovery tools; provides clearer commercial-use guidance than raw license documents, though less authoritative than legal counsel

Maintains a longitudinal view of the open-source LLM ecosystem by tracking model releases, organizational contributions, licensing trends, and parameter-size distributions over time. The repository serves as a historical record of which organizations are releasing open models, when they were released, and how the landscape has evolved, enabling analysis of ecosystem maturity and competitive dynamics.

Unique: Provides a curated, human-reviewed historical record of open-source LLM releases with explicit commercial-use filtering, rather than automated scraping of all models, enabling cleaner trend analysis and reducing noise from research-only or restricted models

vs alternatives: More selective and legally-focused than raw Hugging Face statistics; provides organizational and licensing context that raw model counts lack, though less comprehensive than exhaustive ecosystem surveys

Provides structured information to support model selection decisions by presenting models in a filterable, comparable format with key decision criteria (license, parameter count, organization, release date). The registry enables side-by-side comparison of models and helps developers quickly narrow down options based on their specific constraints (budget, licensing requirements, model size, organizational preference).

Unique: Focuses on commercial-use licensing as a primary decision criterion alongside technical attributes, addressing the specific decision-making needs of enterprises and startups that cannot use restricted models

vs alternatives: More legally-aware than generic model comparison tools; provides clearer filtering for commercial use cases, though less comprehensive than full benchmarking suites that include performance metrics

Aggregates 67 TB of source code from the Software Heritage archive, filtering for permissively licensed repositories (MIT, Apache 2.0, BSD, etc.) across 600+ programming languages. Uses automated license detection and validation to ensure legal compliance for model training. Implements a rigorous deduplication pipeline at file and repository levels to eliminate redundant training data and reduce dataset bloat.

Unique: Largest open-source code dataset at 67 TB with automated opt-out governance allowing repository owners to request removal, combined with rigorous deduplication and PII removal pipeline — no other public dataset offers this scale with legal compliance and community control mechanisms

vs alternatives: Larger and more legally compliant than GitHub's CodeSearchNet (14M files) or Google's BigQuery public datasets, with explicit opt-out governance vs. implicit inclusion, and covers 600+ languages vs. Codex training data's undisclosed language distribution

Implements a community-driven opt-out system where repository owners can request removal of their code from the dataset without legal takedown notices. Maintains a registry of excluded repositories and re-applies exclusions during dataset updates. Provides transparent governance documentation and a clear submission process for removal requests, balancing open access with creator rights.

Unique: First large-scale code dataset to implement opt-out governance at dataset level rather than relying solely on license compliance, with transparent registry and community submission process — shifts power from dataset creators to code contributors

vs alternatives: More respectful of creator autonomy than GitHub Copilot's training approach (no opt-out) or academic datasets (one-time snapshot), and more scalable than individual DMCA takedowns

Automated pipeline that scans source code for personally identifiable information (email addresses, API keys, SSH keys, credit card patterns, phone numbers) and removes or redacts them before dataset release. Uses regex patterns, entropy-based detection for secrets, and heuristic rules to identify sensitive data. Operates at file level with configurable sensitivity thresholds to balance data utility against privacy risk.

Unique: Combines regex pattern matching, entropy-based secret detection, and heuristic rules in a unified pipeline with configurable sensitivity — more comprehensive than simple regex-only approaches, but trades off false positive rate against security coverage

vs alternatives: More thorough than GitHub's secret scanning (which only flags known patterns) because it includes entropy-based detection for unknown secret formats, but less accurate than specialized tools like TruffleHog due to language-agnostic approach

Indexes 67 TB of source code across 600+ programming languages with language-aware metadata (syntax, file extension, language family). Enables retrieval by language, license, repository, or code patterns. Uses Software Heritage's existing indexing infrastructure as foundation, augmented with language detection and classification. Supports both bulk download and filtered queries for specific language subsets.

Unique: Leverages Software Heritage's existing language detection and indexing infrastructure, then augments with BigCode-specific language classification and filtering — avoids reinventing language detection while providing dataset-specific query capabilities

vs alternatives: More comprehensive language coverage (600+ languages) than GitHub's Linguist (500+ languages) and more accessible than Software Heritage's raw API because it's pre-filtered for permissive licenses and deduplicated

Removes duplicate code files and repositories using content hashing (SHA-256 or similar) and fuzzy matching for near-duplicates. Operates in two stages: exact deduplication via hash matching, then fuzzy matching (e.g., Jaccard similarity or MinHash) to catch semantically identical code with minor formatting differences. Preserves one canonical copy of each unique code pattern while removing redundant training examples.

Unique: Two-stage deduplication combining exact hash matching with fuzzy similarity matching (likely MinHash or Jaccard) to catch both identical and near-identical code — more thorough than single-stage approaches but computationally expensive

vs alternatives: More aggressive deduplication than CodeSearchNet (which uses simple hash matching) because it catches near-duplicates, but less semantic than clone detection tools (which understand code structure) because it's content-based

Integrates with Software Heritage's comprehensive archive of 200+ million repositories and their full version control history. Extracts source code snapshots from Software Heritage's Git/Mercurial/SVN repositories, preserving repository metadata (commit history, author info, timestamps). Provides access to code at specific points in time, enabling historical analysis or training on code evolution patterns.

Unique: Leverages Software Heritage's universal code archive (200M+ repositories) as data source, providing access to code that would be impossible to collect via GitHub API alone — enables training on archived/deleted repositories and non-GitHub platforms (GitLab, Gitea, etc.)

vs alternatives: More comprehensive than GitHub-only datasets because it includes code from GitLab, Gitea, SourceForge, and other platforms archived by Software Heritage; more legally defensible than web scraping because it uses an established, community-maintained archive

Tracks and validates SPDX license identifiers for each repository, ensuring only permissively licensed code (MIT, Apache 2.0, BSD, etc.) is included. Maintains license metadata alongside code files, enabling downstream users to verify legal compliance. Implements license hierarchy and compatibility checking to handle dual-licensed or complex licensing scenarios.

Unique: Combines automated SPDX detection with manual review and maintains license metadata alongside code, enabling downstream users to verify compliance — more transparent than datasets that simply claim 'permissive licenses' without proof

vs alternatives: More legally rigorous than GitHub's CodeSearchNet (which doesn't validate licenses) and more transparent than Codex training data (which doesn't disclose license filtering at all)

Maintains versioned snapshots of the dataset (e.g., v2.0, v2.1) with documented changes between versions (new repositories added, deduplication improvements, PII removal updates). Provides checksums and manifests for reproducibility, enabling researchers to cite specific dataset versions and reproduce results. Tracks dataset lineage and transformation history.

Unique: Maintains semantic versioning and detailed changelogs for dataset releases, enabling researchers to cite specific versions and understand dataset evolution — more rigorous than one-off dataset releases without versioning

vs alternatives: More reproducible than academic datasets that are released once without versioning, and more transparent than commercial datasets (Codex) that don't disclose version history or changes