[JS] Implement Text2ImagePipeline

[Retribution98]

Signed-off-by: Kirill Suvorov kirill.suvorov@intel.com

Description

• Implemented Node.js binding for Text2ImagePipeline
• Added Performance metrics for image generation
• Step callback support (except intermediate tensors)
• The simplest sample text2image.js
• Updated hugging_face.py to correct download flux model

CVS-184436

Checklist:

• This PR follows GenAI Contributing guidelines.
• Tests have been updated or added to cover the new code.
• This PR fully addresses the ticket.
• I have made corresponding changes to the documentation.

[Copilot]

Pull request overview

Adds Node.js bindings for OpenVINO GenAI Text2ImagePipeline, including perf metrics exposure, step-callback support, and end-to-end coverage through JS samples/tests and documentation updates. It also adjusts Python test utilities to support exporting/downloading a tiny FLUX model for image-generation testing.

Changes:

• Implemented Text2ImagePipeline Node.js addon wrapper + TypeScript API surface (generate/config/perf metrics + step callback).
• Added JS tests and a minimal JS sample for text-to-image generation, and wired the sample into Python sample tests.
• Updated docs to include JavaScript usage snippets for image generation; updated HF test utility export logic for FLUX.

Reviewed changes

Copilot reviewed 28 out of 29 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
tests/python_tests/utils/hugging_face.py	Adds optional-tokenizer flow + task-specific optimum-cli export override (incl. tiny FLUX).
tests/python_tests/samples/test_text2image.py	Extends sample test to execute the JS text2image sample.
src/js/tests/text2ImagePipeline.test.js	Adds Node.js test coverage for Text2ImagePipeline creation/generate/config/perf/callback behavior.
src/js/tests/setup.py	Adds download/convert setup for an image-generation test model (tiny FLUX).
src/js/src/text2image_pipeline/pipeline_wrapper.cpp	C++ N-API wrapper for init/generate/config/perf metrics + step callback wiring.
src/js/src/text2image_pipeline/perf_metrics.cpp	N-API wrapper for image-generation perf metrics and raw metrics access.
src/js/src/text2image_pipeline/init_worker.cpp	Async init worker for constructing the C++ Text2ImagePipeline off-thread.
src/js/src/helper.cpp	Adds JS↔C++ conversions for ImageGenerationConfig and image-gen perf metrics.
src/js/src/addon.cpp	Registers Text2ImagePipeline and Text2ImagePerfMetrics classes in the addon.
src/js/lib/utils.ts	Adds TS types for ImageGenerationConfig, pipeline properties, and per-step callback.
src/js/lib/pipelines/text2ImagePipeline.ts	Implements the JS/TS Text2ImagePipeline wrapper (promisified init/generate).
src/js/lib/perfMetrics.ts	Adds TS interface/types for Text2Image perf metrics + raw metrics.
src/js/lib/index.ts	Exposes Text2ImagePipeline via the factory/export surface.
src/js/lib/addon.ts	Adds addon-side TS typings for Text2ImagePipeline and Text2ImagePerfMetrics.
src/js/include/text2image_pipeline/pipeline_wrapper.hpp	Declares the N-API wrapper class for Text2ImagePipeline.
src/js/include/text2image_pipeline/perf_metrics.hpp	Declares the N-API wrapper class for Text2Image perf metrics.
src/js/include/text2image_pipeline/init_worker.hpp	Declares the async init worker for Text2ImagePipeline.
src/js/include/helper.hpp	Declares helper template specializations for new image-gen bindings.
src/js/include/addon.hpp	Adds stored prototypes for Text2ImagePipeline and Text2ImagePerfMetrics.
src/js/eslint.config.cjs	Allows snake_case config keys used by image generation options.
site/src/pages/_sections/UseCasesSection/components/image-generation.tsx	Adds JavaScript tab to the image generation use-case section.
site/docs/use-cases/image-generation/_sections/_usage_options/index.mdx	Adds JS snippet for configuring image generation parameters.
site/docs/use-cases/image-generation/_sections/_run_model/index.mdx	Adds JS tab/device examples for running text-to-image.
site/docs/use-cases/image-generation/_sections/_run_model/_text2image_js.mdx	New JS code example (save result via Jimp).
site/docs/bindings/node-js.md	Updates Node.js bindings coverage list to include Text2ImagePipeline.
samples/js/package.json	Adds jimp for the image-generation JS sample.
samples/js/package-lock.json	Lockfile update to include Jimp and transitive deps.
samples/js/image_generation/text2image.js	New minimal JS Text2Image sample with per-step callback + BMP output.
samples/js/image_generation/README.md	New README documenting the JS text-to-image sample usage.

Files not reviewed (1)

• samples/js/package-lock.json: Language not supported

[Copilot]

Pull request overview

Copilot reviewed 29 out of 30 changed files in this pull request and generated 7 comments.

Files not reviewed (1)

• samples/js/package-lock.json: Language not supported

Comments suppressed due to low confidence (1)

tests/python_tests/utils/hugging_face.py:1

• get_huggingface_models() can now return hf_tokenizer=None, but the provided context shows OVConvertedModelSchema.hf_tokenizer is typed as AutoTokenizer (non-optional). This mismatch can lead to downstream code assuming a tokenizer exists and failing at runtime or during type checking; update the schema annotation (and any consumers) to accept AutoTokenizer | None, or ensure the schema is only constructed when a tokenizer is present.

# Copyright (C) 2018-2026 Intel Corporation