{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreiaa6f5ssg24afxopltufkapts3i4mlsdbvjj52cmnnahesmbemvgu",
"uri": "at://did:plc:nfto3lv2rcs5s7h7digotzlu/app.bsky.feed.post/3mngjnylp62b2"
},
"coverImage": {
"$type": "blob",
"ref": {
"$link": "bafkreig7saeenw6c3prrms24e2bpcmq7mskq26az4ajg73ufx33cf4gf2a"
},
"mimeType": "image/png",
"size": 32236
},
"path": "/packages/model2vec",
"publishedAt": "2026-06-04T02:00:17.581Z",
"site": "https://pub.dev",
"textContent": "A high-performance Dart wrapper for model2vec-rs using Rust FFI. Generate fast, local, and static text embeddings with minimal memory footprint using Native Assets. Changelog excerpt: **New Features:** - `getRecommendedModels()`no longer calls FFI — now returns a hardcoded list of 7 models - Removed `get_model_list`from FFI bindings (Rust, Dart, `.h`) - `generateEmbedding()`now accepts `maxLength`parameter — signature changed - `generateBatchEmbeddings()`now accepts `maxLength`and `batchSize`parameters — signature changed - **Streaming API**— `generateEmbeddingStream()`for processing large datasets with batching and optional worker isolate - **Async API**— `g[...]",
"title": "v1.1.0 of model2vec",
"updatedAt": "2026-06-04T01:58:13.414Z"
}