This release adds 1 notable feature for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Topics
+6 more
Summary
AI summaryVisual ingest adds optional 'quality' profile for improved capture of in‑image text.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Medium |
Visual ingest now offers two quality profiles (opt-in). Visual ingest now offers two quality profiles (opt-in). Source: granite4.1:8b-q6_K@2026-05-21 Confidence: low |
— |
| Feature | Low |
Visual ingest supports 'fast' and 'quality' quality profiles. Visual ingest supports 'fast' and 'quality' quality profiles. Source: granite4.1:30b@2026-05-21-audit Confidence: low |
— |
| Feature | Low |
'fast' profile uses HuggingFaceTB/SmolVLM-256M-Instruct model by default. 'fast' profile uses HuggingFaceTB/SmolVLM-256M-Instruct model by default. Source: granite4.1:30b@2026-05-21-audit Confidence: low |
— |
| Feature | Low |
'quality' profile uses onnx-community/Qwen2.5-VL-3B-Instruct-ONNX for better text capture. 'quality' profile uses onnx-community/Qwen2.5-VL-3B-Instruct-ONNX for better text capture. Source: granite4.1:30b@2026-05-21-audit Confidence: low |
— |
Full changelog
Changes
- Visual ingest now offers two quality profiles (opt-in). When
visual: true(MCP) or--visual(CLI) is set, a newvisualQualityparameter (visualQuality: 'fast' | 'quality'on MCPingest_file;--visual-quality fast|qualityon CLI) selects the VLM.fast(default) keeps the v0.14.0 modelHuggingFaceTB/SmolVLM-256M-Instruct.qualityselectsonnx-community/Qwen2.5-VL-3B-Instruct-ONNXfor figures where in-image text (axis labels, panel sub-labels, flowchart nodes) needs to be captured reliably. Default behavior is unchanged; thequalityprofile is opt-in per ingest call.
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About shinpr/mcp-local-rag
Privacy-first document search server running entirely locally. Supports semantic search over PDFs, DOCX, TXT, and Markdown files with LanceDB vector storage and local embeddings - no API keys or cloud services required.
Related context
Beta — feedback welcome: [email protected]