Documented fields
How many records have a verified or classified field?
Explore license evidence →Explore 811 field-verified AI model records. Understand licenses, compare checkpoints and trace what the original sources actually say.
GO FROM UPDATES TO DECISIONS
Three first-party research paths built from our reviewed source records.
Source-grounded fit, limitations and side-by-side model evidence.
Explore model paths →02 · VERIFYModel Change IntelligenceObserved weight, access, architecture and license-field revisions.
Review source changes →03 · STAY INFORMEDWeekly intelligenceA concise index of our genuinely published, primary-source News.
Read the briefing →NEW · OPEN-WEIGHT AGENT INTELLIGENCE
Explore 85 Leading and Notable open-weight models by their role in agent systems. Compare reasoning cores, retrieval components and media tools with traceable license, context and hardware evidence.
Explore models for AI agents →OPEN WEIGHT INTELLIGENCE
Original source-linked reporting, with a direct route to model evidence. Browse all News →
A public LoRA and merged checkpoint adapt a huge sparse model for software agents; publisher benchmarks suggest a narrow gain, not universal superiority.
Read the evidence briefThe new 224B multimodal model promises cheaper agent workflows; Tencent has opened commercial API access, not a verifiable public weight release.
Read the evidence briefThe Apache-2.0 base checkpoint listens and speaks simultaneously, with openly available local inference code and carefully qualified latency claims.
Read the evidence briefHugging Face expands native open-weight architecture support and introduces an ExportedModel interface for moving compatible checkpoints into supported inference runtimes.
Read the evidence briefTHREE WAYS TO USE OPEN MODEL WEIGHTS
Start with a model, inspect its evidence, then follow the changes that matter. No composite rankings or invented compatibility claims.
Search 811 source-verified records. Filter by license, context, published formats and weight-only memory estimates.
Search the model registry →Inspect original sources, exact repository revisions, field evidence and dated changes before drawing conclusions.
Explore verified evidence →Review changes to models saved in this browser. Email and background alerts are not available yet.
Open the monitoring preview →MODEL-SPECIFIC EVIDENCE ANSWERS
Read a dated source comparison before inferring a model or license change. These are observed examples, not a model ranking.
A source repository SHA changed. Read which fields the retained comparison actually documents.
Read the dated evidence →Observed repository revision · 2026-10-11Wan2.1A source repository SHA changed. Read which fields the retained comparison actually documents.
Read the dated evidence →OMW evidence refinement · 2026-10-07Qwen3-32BThe source SHA stayed unchanged. See how OMW evidence classification changed without a claimed publisher release.
Read the dated evidence →Need to compare more than once? Save models in a private shortlist → Your selection stays in this browser.
SOURCE-VERIFIED · NO MODEL RANKINGS
Counts from 811 verified model records. Not intelligence, speed or quality scores.
How many records have a verified or classified field?
Explore license evidence →Current claim-schema refresh progress across the registry.
Missing source claims stay unknown, never false.
Inspect source comparisons →Repository-observed technical evidence, not runtime tests.
A runtime signal does not prove successful execution.
Review verification scope →Bars show field coverage as a share of the verified registry. They do not compare model capabilities. Read the methodology →
TEMPORAL EVIDENCE GRAPH
Lists are easy to copy. A dated record of what was observed, where it came from and how it changed is harder to recreate.
Observed field snapshots, source revisions, evidence hashes and material diffs accumulate instead of being overwritten.
Explore the evidence graph →COMPATIBILITYConnect runtimes, formats, precisions, licenses, base models and memory constraints.
Open compatibility graph →FRESHNESSRepository revisions are checked every day; changed and new sources return to full field verification.
See recent changes →MACHINE LAYERVersioned JSON, OpenAPI and a read-only MCP endpoint expose the same evidence as the website.
Connect through MCP →VERIFIED RECORDS
262,144 tokens · nvidia-nemotron-open-model-license
1,048,576 tokens · Apache-2.0
40,960 tokens · Apache-2.0
163,840 tokens · MIT
131,072 tokens · Apache-2.0
128,000 tokens · gemma
STAY CONNECTED
Choose how to stay up to date. No account or email subscription required.
Discover model news, useful comparisons and important developments, with links back to the evidence.
Read updates in your favorite RSS app. Follow the news, model-source changes, or both.
Save the website to your phone’s Home Screen or bookmark it on your computer.
iPhone / iPad (Safari): Tap Share, then Add to Home Screen.
Android (Chrome): Open the browser menu (⋮), then choose Add to Home screen or Install app, when offered.
Desktop: Bookmark this page with ⌘ + D (Mac) or Ctrl + D (Windows / Linux).
Following individual models? Open your personal watchlist →
PRACTICE GUIDES · REGISTRY-BACKED
Explore hardware decisions, source transparency and local coding workflows using our verified model records. Estimates remain estimates, and unknown facts stay unknown.
SOURCE-FIRST BY DESIGN
Open Model Weights separates repository evidence, publisher documentation and derived values. Unknown stays unknown, and popularity signals never become a quality score.