OVMS 2026.0.0.9944c3235 - Unsupported 'qwen3_vl' VLM model type

[Arol450]

Trying to use OVMS latest weekly build to serve Qwen3-VL-8B-Instruct/INT8 model.
OVMS 2026.0.0.9944c3235
Model was exported using instructions from here:
https://github.com/openvinotoolkit/openvino_notebooks/tree/latest/notebooks/qwen3-vl
Accumulated export file size is about 8+G.

To reproduce start serving said model with said OVMS,
--source_model "Qwen3-VL-8B-Instruct/INT8" --model_repository_path /models --rest_port 11434 --target_device GPU.0 --task text_generation
resulting in the following relevant error messages:

[1][serving][info][servablemanagermodule.cpp:51] ServableManagerModule starting
[1][serving][info][mediapipegraphdefinition.cpp:423] MediapipeGraphDefinition initializing graph nodes
[1][modelmanager][info][servable_initializer.cpp:431] Initializing Visual Language Model Continuous Batching servable
[serving][error][servable_initializer.cpp:214] Error during llm node initialization for models_path: /models/Qwen3-VL-8B-Instruct/INT8/./ exception: Exception from ../../../../../repos/openvino.genai/src/cpp/src/visual_language/vlm_config.cpp:33:
Unsupported 'qwen3_vl' VLM model type
[modelmanager][error][servable_initializer.cpp:436] Error during LLM node resources initialization: The LLM Node resource initialization failed
[1][serving][error][mediapipegraphdefinition.cpp:474] Failed to process LLM node graph Qwen3-VL-8B-Instruct/INT8
[1][modelmanager][info][pipelinedefinitionstatus.hpp:59] Mediapipe: Qwen3-VL-8B-Instruct/INT8 state changed to: LOADING_PRECONDITION_FAILED after handling: ValidationFailedEvent:
[1][modelmanager][error][modelmanager.cpp:184] Couldn't start model manager

I assumed 'qwen3_vl' VLM model type would be supported starting 2026 version.
Is this not the case?
Any ideas as to when support is expected, or is this just a bug?