Skip to content

v0.4.0 — Qwen 3 1.7B Default Model, Multi-Model Support

Choose a tag to compare

View all tags
@pluginslab pluginslab released this 07 Mar 13:04
· 378 commits to main since this release
v0.4.0
6777eb5

What's New

Qwen 3 1.7B as Default Model

  • ~3x faster inference compared to Qwen 2.5 7B
  • ~1.2GB download (down from ~4.5GB)
  • 93% E2E pass rate (14/15 tests) with 100% JSON reliability
  • Native function calling support — no JSON parsing fallback needed

Multi-Model Support

  • Users can now choose between Qwen 3 1.7B (default, fast) and Qwen 2.5 7B (alternative, advanced reasoning)
  • Model preference persisted in localStorage across sessions
  • Both models work in Service Worker persistent mode

E2E Test Results

Model Pass Rate JSON Reliability Download Size Strengths
Qwen 3 1.7B (default) 93% (14/15) 100% ~1.2GB Fast inference, native tool calling, lightweight
Qwen 2.5 7B 96% (26/27) 100% ~4.5GB Multi-step reasoning, complex workflows

Files Changed

  • model-loader.js — new default model, getAvailableModels() returns both options
  • README.md, ARCHITECTURE.md — updated documentation for dual-model support
  • tests/e2e/RESULTS.md — added Qwen 3 1.7B detailed test results
  • tests/e2e/config.js — default test model updated

Full Changelog: v0.3.1...v0.4.0

Assets 2
Loading