Add Qwen 3.5, Qwen 3.6, and MoonshotAI Kimi workflow blocks via OpenRouter (#2280)

* Add Qwen 3.5, Qwen 3.6, and MoonshotAI Kimi workflow blocks via OpenRouter

Three new VLM workflow blocks served through OpenRouter, mirroring the
existing google_gemma block. Each supports the standard task surface
(unconstrained, ocr, structured-answering, classification, multi-label,
VQA, caption, detailed-caption, object-detection) and ships with unit
tests covering manifest validation and the request execution path.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* Run isort on loader.py

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* Update Google Gemma -> Google Gemma API, so it's clear its via API. This will allow us to add "selfhosted" gemma 4 later

---------

Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Co-authored-by: Paweł Pęczek <146137186+PawelPeczek-Roboflow@users.noreply.github.com>

Author: 3 people

inference/core/workflows/core_steps/loader.py

@@ -248,6 +248,9 @@
 from inference.core.workflows.core_steps.models.foundation.google_vision_ocr.v1 import (
     GoogleVisionOCRBlockV1,
 )
+from inference.core.workflows.core_steps.models.foundation.kimi_openrouter.v1 import (
+    KimiOpenRouterBlockV1,
+)
 from inference.core.workflows.core_steps.models.foundation.llama_vision.v1 import (
     LlamaVisionBlockV1,
 )
@@ -274,9 +277,15 @@
 from inference.core.workflows.core_steps.models.foundation.perception_encoder.v1 import (
     PerceptionEncoderModelBlockV1,
 )
+from inference.core.workflows.core_steps.models.foundation.qwen3_5_openrouter.v1 import (
+    Qwen35OpenRouterBlockV1,
+)
 from inference.core.workflows.core_steps.models.foundation.qwen3_5vl.v1 import (
     Qwen35VLBlockV1,
 )
+from inference.core.workflows.core_steps.models.foundation.qwen3_6_openrouter.v1 import (
+    Qwen36OpenRouterBlockV1,
+)
 from inference.core.workflows.core_steps.models.foundation.qwen3vl.v1 import (
     Qwen3VLBlockV1,
 )
@@ -918,6 +927,9 @@ def load_blocks() -> List[Type[WorkflowBlock]]:
         Qwen25VLBlockV1,
         Qwen3VLBlockV1,
         Qwen35VLBlockV1,
+        Qwen35OpenRouterBlockV1,
+        Qwen36OpenRouterBlockV1,
+        KimiOpenRouterBlockV1,
         SmolVLM2BlockV1,
         Moondream2BlockV1,
         OverlapBlockV1,

inference/core/workflows/core_steps/models/foundation/google_gemma/v1.py

@@ -114,9 +114,9 @@
 class BlockManifest(WorkflowBlockManifest):
     model_config = ConfigDict(
         json_schema_extra={
-            "name": "Google Gemma",
+            "name": "Google Gemma API",
             "version": "v1",
-            "short_description": "Run Google's Gemma model with vision capabilities.",
+            "short_description": "Run Google's Gemma model with vision capabilities via OpenRouter.",
             "long_description": LONG_DESCRIPTION,
             "license": "Gemma Terms of Use",
             "block_type": "model",