Skip to content

FlowSight local weights v0.3.0 (Qwen3-VL-2B-Instruct)

Pre-release
Pre-release

Choose a tag to compare

View all tags
@Mancasvel Mancasvel released this 29 Sep 16:48
· 16 commits to main since this release
models-v0.3.0
3a426f6
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

FlowSight local model assets v0.3.0

This asset release supplies the single local model for FlowSight 5.0.3: the official Qwen3-VL-2B-Instruct GGUF from the Qwen team. It is data for on-device inference, not an application installer or a FlowSight fine-tune. The model is licensed under Apache License 2.0; see the bundled model notice and license in the FlowSight application.

Upstream GGUF repository: Qwen/Qwen3-VL-2B-Instruct-GGUF, revision 52d6c8ffea26cc873ac5ad116f8631268d7eb503.

Asset Size (bytes) SHA-256
Qwen3VL-2B-Instruct-Q4_K_M.gguf 1,107,409,952 089d75c52f4b7ffc56ba998ffc50aae89fcafc755f9e7208aacca281dca6c2ae
mmproj-Qwen3VL-2B-Instruct-Q8_0.gguf 445,053,216 f9a68fabba69c3b81e153367b2c7521030b0fa8bb0de400c9599c8e6725f9c82

FlowSight verifies both size and SHA-256 before using either file. Windows and macOS download these assets on first use; the Linux installer includes them. No screen content is sent to the model host.

Assets 4
Loading