{"id":13152,"date":"2026-11-17T11:02:09","date_gmt":"2026-11-17T02:02:09","guid":{"rendered":"https:\/\/localmodelwatch.tsuchitsuchi.com\/?p=13152"},"modified":"2026-11-17T11:11:17","modified_gmt":"2026-11-17T02:11:17","slug":"comfy-org-ming-3-0-tts-2","status":"publish","type":"post","link":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/comfy-org-ming-3-0-tts-2\/","title":{"rendered":"Ming-3.0-TTS Multilingual TTS Model: 24GB+ VRAM, ComfyUI Paths"},"content":{"rendered":"<p><!-- lmw:facts --><\/p>\n<h2>At a Glance<\/h2>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Item<\/th>\n<th>Value<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Repository<\/td>\n<td><a href=\"https:\/\/huggingface.co\/Comfy-Org\/Ming-3.0-TTS\">Comfy-Org\/Ming-3.0-TTS<\/a><\/td>\n<\/tr>\n<tr>\n<td>Publisher guide<\/td>\n<td><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/publisher-inclusionai-en\/\">inclusionAI (Ant Group): models and licenses<\/a><\/td>\n<\/tr>\n<tr>\n<td>Published<\/td>\n<td>2026-10-23<\/td>\n<\/tr>\n<tr>\n<td>License<\/td>\n<td>mit<\/td>\n<\/tr>\n<tr>\n<td>Formats<\/td>\n<td><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/format-safetensors-en\/\">safetensors<\/a><\/td>\n<\/tr>\n<tr>\n<td>Source type<\/td>\n<td>Primary source (the publisher itself)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>Values determined by this site&#8217;s code when the information was collected. Dates are JST.<\/em><\/p>\n<p><!-- \/lmw:facts --><\/p>\n<h2>Overview<\/h2>\n<p>Comfy-Org has released the open-weight text-to-speech model &#8220;Comfy-Org\/Ming-3.0-TTS&#8221;. This model is a checkpoint repackaged for ComfyUI based on the speech synthesis foundation model &#8220;Ming-3.0-tts-tiny&#8221; developed by inclusionAI. By inputting text information and reference audio, it can generate multilingual and expressive speech.<\/p>\n<h2>Specifications<\/h2>\n<p>The specifications of this model, based on the original model &#8220;inclusionAI\/Ming-3.0-tts-tiny&#8221; and the distribution repository, are as follows:<\/p>\n<ul>\n<li>Parameter Count: Total 7.9B parameters, activating approximately 1.3B parameters per token (7.9B-A1.3B)<\/li>\n<li>Architecture: LingV3TTSForConditionalGeneration (lingv3_tts, Mixture of Experts)<\/li>\n<li>Output Specifications: Mono 44.1 kHz WAV (example when using vLLM-Omni)<\/li>\n<li>Supported Languages: Chinese, English, German, Spanish, French, Korean, Japanese, and Chinese dialects (Sichuan dialect, Cantonese)<\/li>\n<li>Recommended Inference Settings: cfg_scale=2.0, temperature=0, flow_steps=10, sigma=0.25 (Python API default values of the original model)<\/li>\n<li>License Terms: MIT license applies, permitting use including commercial use as long as the prescribed copyright notice is retained<\/li>\n<\/ul>\n<h2>Performance and Quality<\/h2>\n<p>Detailed comparative benchmark data regarding performance has been published by the developers of the original model, &#8220;inclusionAI\/Ming-3.0-tts-tiny&#8221;.<\/p>\n<p>Comparisons with various models on the zero-shot speech synthesis evaluation set (Seed-TTS testset) are as follows. Note that lower WER (Word Error Rate) indicates better performance.<\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<p class=\"lmw-table-hint\" style=\"margin:0 0 4px;font-size:0.85em;opacity:0.7;\">\u2192 Scroll horizontally to see all columns<\/p>\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Model<\/th>\n<th>seed-tts-eval-zh \/ WER \u2193<\/th>\n<th>seed-tts-eval-zh \/ SIM \u2191<\/th>\n<th>seed-tts-eval-en \/ WER \u2193<\/th>\n<th>seed-tts-eval-en \/ SIM \u2191<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Seed-TTS<\/td>\n<td>1.110%<\/td>\n<td>0.796<\/td>\n<td>2.240%<\/td>\n<td>0.762<\/td>\n<\/tr>\n<tr>\n<td>GLM-TTS<\/td>\n<td>1.030%<\/td>\n<td>0.761<\/td>\n<td>2.230%<\/td>\n<td>0.672<\/td>\n<\/tr>\n<tr>\n<td>Qwen3-TTS<\/td>\n<td>1.070%<\/td>\n<td>\u2014<\/td>\n<td>1.390%<\/td>\n<td>\u2014<\/td>\n<\/tr>\n<tr>\n<td>Qwen3-TTS-12Hz-0.6B-Base<\/td>\n<td>0.920%<\/td>\n<td>\u2014<\/td>\n<td>1.320%<\/td>\n<td>\u2014<\/td>\n<\/tr>\n<tr>\n<td>Qwen3-TTS-12Hz-1.7B-Base<\/td>\n<td>0.770%<\/td>\n<td>\u2014<\/td>\n<td>1.240%<\/td>\n<td>\u2014<\/td>\n<\/tr>\n<tr>\n<td>CosyVoice3-0.5B<\/td>\n<td>1.160%<\/td>\n<td>0.780<\/td>\n<td>2.020%<\/td>\n<td>0.718<\/td>\n<\/tr>\n<tr>\n<td>CosyVoice3-1.5B<\/td>\n<td>0.710%<\/td>\n<td>0.775<\/td>\n<td>1.450%<\/td>\n<td>0.695<\/td>\n<\/tr>\n<tr>\n<td>F5-TTS<\/td>\n<td>1.560%<\/td>\n<td>0.741<\/td>\n<td>1.830%<\/td>\n<td>0.647<\/td>\n<\/tr>\n<tr>\n<td>Ming-3.0-tts-0.5b<\/td>\n<td>0.800%<\/td>\n<td>0.758<\/td>\n<td>1.584%<\/td>\n<td>0.691<\/td>\n<\/tr>\n<tr>\n<td>Ming-3.0-tts-tiny<\/td>\n<td>0.820%<\/td>\n<td>0.781<\/td>\n<td>1.666%<\/td>\n<td>0.728<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>According to the figures published by the release source, Ming-3.0-tts-tiny recorded a Chinese WER of 0.820%, achieving an accuracy that surpasses models such as GLM-TTS (1.030%) and CosyVoice3-0.5B (1.160%). It also showed top-tier results with a Chinese Speaker Similarity (SIM) of 0.781. On the other hand, its English WER is 1.666%, falling short of some models such as the Qwen3-TTS series and CosyVoice3-1.5B.<\/p>\n<p>Next, the evaluation results in a multilingual environment (MiniMax test set) are as follows. Lower CER (Character Error Rate) and WER indicate better performance.<\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<p class=\"lmw-table-hint\" style=\"margin:0 0 4px;font-size:0.85em;opacity:0.7;\">\u2192 Scroll horizontally to see all columns<\/p>\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Model<\/th>\n<th>German \/ WER \u2193<\/th>\n<th>German \/ SIM \u2191<\/th>\n<th>Spanish \/ WER \u2193<\/th>\n<th>Spanish \/ SIM \u2191<\/th>\n<th>French \/ WER \u2193<\/th>\n<th>French \/ SIM \u2191<\/th>\n<th>Korean \/ CER \u2193<\/th>\n<th>Korean \/ SIM \u2191<\/th>\n<th>Japanese \/ CER \u2193<\/th>\n<th>Japanese \/ SIM \u2191<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>dots.tts<\/td>\n<td>0.452%<\/td>\n<td>0.798<\/td>\n<td>2.593%<\/td>\n<td>0.839<\/td>\n<td>19.878%<\/td>\n<td>0.789<\/td>\n<td>4.837%<\/td>\n<td>0.853<\/td>\n<td>5.134%<\/td>\n<td>0.838<\/td>\n<\/tr>\n<tr>\n<td>VoxCPM2<\/td>\n<td>0.963%<\/td>\n<td>0.790<\/td>\n<td>1.324%<\/td>\n<td>0.836<\/td>\n<td>3.380%<\/td>\n<td>0.761<\/td>\n<td>3.547%<\/td>\n<td>0.839<\/td>\n<td>4.358%<\/td>\n<td>0.825<\/td>\n<\/tr>\n<tr>\n<td>Fish Audio S2<\/td>\n<td>1.045%<\/td>\n<td>0.714<\/td>\n<td>1.216%<\/td>\n<td>0.748<\/td>\n<td>3.636%<\/td>\n<td>0.698<\/td>\n<td>1.675%<\/td>\n<td>0.761<\/td>\n<td>3.427%<\/td>\n<td>0.755<\/td>\n<\/tr>\n<tr>\n<td>FireRedTTS3<\/td>\n<td>0.348%<\/td>\n<td>0.825<\/td>\n<td>1.086%<\/td>\n<td>0.868<\/td>\n<td>4.366%<\/td>\n<td>0.807<\/td>\n<td>1.644%<\/td>\n<td>0.864<\/td>\n<td>3.418%<\/td>\n<td>0.825<\/td>\n<\/tr>\n<tr>\n<td>Ming-3.0-tts-0.5b<\/td>\n<td>0.365%<\/td>\n<td>0.753<\/td>\n<td>1.099%<\/td>\n<td>0.763<\/td>\n<td>4.264%<\/td>\n<td>0.675<\/td>\n<td>1.728%<\/td>\n<td>0.793<\/td>\n<td>3.765%<\/td>\n<td>0.765<\/td>\n<\/tr>\n<tr>\n<td>Ming-3.0-tts-tiny<\/td>\n<td>0.325%<\/td>\n<td>0.791<\/td>\n<td>0.971%<\/td>\n<td>0.797<\/td>\n<td>3.470%<\/td>\n<td>0.733<\/td>\n<td>6.000%<\/td>\n<td>0.797<\/td>\n<td>3.427%<\/td>\n<td>0.780<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>In measurements by the release source, the German WER (0.325%) and Spanish WER (0.971%) are the best among the compared models. The Japanese CER also recorded a favorable level at 3.427%. On the other hand, the Korean CER is 6.000%, which is higher compared to other models, and the model card from the release source points out the accuracy discrepancy across languages as an issue.<\/p>\n<h2>Strengths and Use Cases<\/h2>\n<p>&#8220;Ming-3.0-tts-tiny&#8221;, which serves as the base for this model, is characterized not only by reading text aloud but also by possessing a variety of control functions and flexible speech expression capabilities. This repository, &#8220;Comfy-Org\/Ming-3.0-TTS&#8221;, repackages these powerful features so they can be utilized in ComfyUI&#8217;s node-based environment.<\/p>\n<p>Based on documentation for the original model, it is particularly adept at the following use cases and controls:<\/p>\n<h3>Zero-Shot Voice Cloning and Natural Language Voice Design<\/h3>\n<p>It supports zero-shot voice cloning, where providing a short reference audio clip allows the model to reproduce that speaker&#8217;s voice quality and speak. Furthermore, even if audio material cannot be prepared, it is possible to newly create a target voice timbre from natural language text prompts (Voice Design) such as &#8220;A young woman&#8217;s voice, clear and natural&#8221;. In addition, over 100 built-in Chinese voices and 5 built-in English voices are provided in advance, allowing for easy synthesis without specifying reference audio.<\/p>\n<h3>Inline Control of Diverse Para-linguistic Expressions, Emotions, and Speaking Speed<\/h3>\n<p>By embedding inline tags within the text, 14 types of non-verbal and para-linguistic expressions\u2014such as breathing, coughing, sighing, laughter, hesitation, surprise, and questioning\/confirmation\u2014can be reproduced in Chinese and English. It also features functions to explicitly adjust parameters such as emotion specification, speaking speed, pitch, and volume. It supports tone control via free-form prompts, enabling subtle nuance adjustments tailored to the context.<\/p>\n<h3>Multilingual Deployment and Dialect Support<\/h3>\n<p>In addition to Chinese and English as main languages, it supports speech synthesis in multiple languages such as German, Spanish, French, Korean, and Japanese. Furthermore, it supports control of specific regional dialects such as the Sichuan dialect and Cantonese alongside standard Chinese, making it applicable to content creation across a wide range of regions and cultural spheres.<\/p>\n<h3>Multi-Speaker Podcast-Style Dialogue Generation<\/h3>\n<p>By preparing reference audio and scripts for multiple speakers, it also supports tasks that batch-synthesize podcast-style dialogue audio where multiple speakers converse back and forth.<\/p>\n<p>Integrating the model into the ComfyUI environment makes it easy to directly link image generation or video generation workflows with this speech synthesis model on the same interface. For example, it can be utilized to build advanced automated pipelines, such as generating audio matching the appearance of a character simultaneously with its image generation, or generating narration tailored to an animated video.<\/p>\n<p><!-- lmw:hardware --><\/p>\n<h2>Hardware Requirements<\/h2>\n<p><strong>Estimated requirements (calculated by Local Model Watch)<\/strong> \u2014 9.0B parameters (taken from the base model inclusionAI\/Ming-3.0-tts-tiny)<\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>Your VRAM<\/th>\n<th>Quantization<\/th>\n<th>File size<\/th>\n<th>Est. memory needed<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>24GB (RTX 4090 \/ 3090, etc.)<\/td>\n<td>BF16<\/td>\n<td>16.7GB<\/td>\n<td>20.1GB<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>Memory estimates add a 20% runtime overhead (KV cache, etc.) to the actual size of the distributed files. Actual usage varies with context length, batch size and inference engine. These figures are computed by this site from file sizes, not published by the model&#8217;s authors. Compare with other models in our <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/vram-guide-en\/\">VRAM quick reference<\/a>. What the quantization names mean: <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/glossary-quantization-en\/\">glossary<\/a>.<\/em><\/p>\n<p><!-- \/lmw:hardware --><\/p>\n<p><!-- lmw:runnability --><\/p>\n<h2>Can You Run It Locally?<\/h2>\n<p>The publisher distributes this model as safetensors.<\/p>\n<p><strong>License \u2014 <code>mit<\/code> (Commercial use allowed):<\/strong> Permits commercial use, modification and redistribution, provided the copyright notice and license text are retained.<\/p>\n<p><em>Compiled by this site&#8217;s code from the published formats and the license field. License summaries are not legal advice \u2014 check the publisher&#8217;s original terms before relying on them.<\/em><\/p>\n<p><!-- \/lmw:runnability --><\/p>\n<p><!-- lmw:files --><\/p>\n<h2>Distributed Files<\/h2>\n<p><em>Weight files published in <a href=\"https:\/\/huggingface.co\/Comfy-Org\/Ming-3.0-TTS\/tree\/main\">Comfy-Org\/Ming-3.0-TTS<\/a>, listed by this site from the Hugging Face API. Sizes are the actual file sizes.<\/em><\/p>\n<div class=\"lmw-table-scroll\" tabindex=\"0\" style=\"overflow-x:auto;-webkit-overflow-scrolling:touch;max-width:100%;\">\n<table style=\"width:max-content;min-width:100%;border-collapse:collapse;\">\n<thead>\n<tr>\n<th>File<\/th>\n<th>Size<\/th>\n<th>ComfyUI folder<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>checkpoints\/ming_tts_tiny_int8_convrot.safetensors<\/code><\/td>\n<td>9.07GB<\/td>\n<td><code>models\/checkpoints\/<\/code><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><em>The ComfyUI folder is shown only where the repository&#8217;s layout mirrors ComfyUI&#8217;s <code>models\/<\/code> directory.<\/em><\/p>\n<p><!-- \/lmw:files --><\/p>\n<h2>How to Get It<\/h2>\n<p>This model is available in the &#8220;Comfy-Org\/Ming-3.0-TTS&#8221; repository on Hugging Face. Since it is not a gated model, you can download the files directly without individually agreeing to terms of use.<\/p>\n<h3>Distribution Format and Placement<\/h3>\n<p>The weight files are distributed as a single safetensors format (<code>ming_tts_tiny_int8_convrot.safetensors<\/code>). When using it in ComfyUI, download the files from the repository and place them according to the following folder structure within your ComfyUI directory:<\/p>\n<pre><code>\ud83d\udcc2 ComfyUI\/\n\u251c\u2500\u2500 \ud83d\udcc2 models\/\n\u2502   \u251c\u2500\u2500 \ud83d\udcc2 checkpoints\/\n\u2502   \u2502   \u2514\u2500\u2500 ming_tts_tiny_int8_convrot.safetensors\n<\/code><\/pre>\n<h3>Inference in Original Model Environments and Servers<\/h3>\n<p>If you perform inference using the standalone code or server features of the original model &#8220;inclusionAI\/Ming-3.0-tts-tiny&#8221; without going through ComfyUI, it is recommended to refer to the GitHub repository (<code>inclusionAI\/Ming-3.0-tts<\/code>) published by the developers. The official repository of the original model provides a verified environment specifying Python 3.10.15, PyTorch 2.7.1+cu118, and transformers 4.46.3.<\/p>\n<p>Additionally, the original model supports provision via OpenAI-compatible endpoints (<code>\/v1\/audio\/speech<\/code>) using vLLM-Omni (assuming version 0.32.0) or SGLang-Omni. By launching a server using these inference engines, external applications can request and receive WAV, MP3, or FLAC audio via a REST API.<\/p>\n<p>Note that regarding voice cloning and the use of built-in voices, the release source explicitly outlines disclaimers prohibiting impersonation of real people without consent, fraud, and illegal purposes. In scenarios where generated audio risks being mistaken for an authentic recording, users are required to clearly state that it is synthesized speech.<\/p>\n<p><!-- lmw:related --><\/p>\n<h2>Related Articles<\/h2>\n<ul>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/comfy-org-releases-world-model-pan2-for-comfyui\/\">PAN2 World Model for Simulating Physical Changes: ComfyUI Paths<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/comfy-org-minimax-h3-image-released\/\">MiniMax-H3-Image 16-Megapixel Image Generation Model: 4GB+ VRAM<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/comfy-org-idol-comfyui-release\/\">Idol 3D Gaussian Splatting Generation Model: ComfyUI Paths<\/a><\/li>\n<li><a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/2026\/11\/17\/comfy-org-fibo-2-0-comfyui\/\">Fibo-2.0 Image Generation and Editing Diffusion Model: 32GB+ VRAM<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:related --><\/p>\n<p><!-- lmw:next-steps --><\/p>\n<h2>What to Read Next<\/h2>\n<ul>\n<li><strong>Find models by VRAM<\/strong> (This model runs from the 24GB tier) \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/vram-guide-24gb-en\/\">Other models that run on a 24GB GPU<\/a><\/li>\n<li><strong>Engines that run this model<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/engine-comfyui-en\/\">ComfyUI<\/a><\/li>\n<li><strong>Formats this model is available in<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/format-safetensors-en\/\">Safetensors format guide and models<\/a><\/li>\n<li><strong>Learn about the publisher<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/publisher-inclusionai-en\/\">inclusionAI (Ant Group): models, licenses and articles<\/a><\/li>\n<li><strong>How to read WER, CER<\/strong> \u2192 <a href=\"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/glossary-benchmarks-en\/\">Benchmark glossary<\/a><\/li>\n<\/ul>\n<p><!-- \/lmw:next-steps --><\/p>\n<h2>Sources<\/h2>\n<ul>\n<li><a href=\"https:\/\/huggingface.co\/Comfy-Org\/Ming-3.0-TTS\">Comfy-Org\/Ming-3.0-TTS (Hugging Face)<\/a><\/li>\n<li><a href=\"https:\/\/huggingface.co\/inclusionAI\/Ming-3.0-tts-tiny\">inclusionAI\/Ming-3.0-tts-tiny (Hugging Face)<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Comfy-Org has released Ming-3.0-TTS, an open-weight text-to-speech model packaged for ComfyUI based on inclusionAI\/Ming-3.0-tts-tiny.<\/p>\n","protected":false},"author":1,"featured_media":13151,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[203],"tags":[1142,677,4048,4050,165,1547,213],"class_list":["post-13152","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-image-video-and-audio","tag-comfy-org-en","tag-comfyui-en","tag-ming-3-0-tts-en","tag-ming-3-0-tts-tiny-en","tag-moe-en","tag-verified","tag--en"],"lang":"en","translations":{"en":13152,"ja":13150},"_links":{"self":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/13152","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/comments?post=13152"}],"version-history":[{"count":2,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/13152\/revisions"}],"predecessor-version":[{"id":17921,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/posts\/13152\/revisions\/17921"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media\/13151"}],"wp:attachment":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media?parent=13152"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/categories?post=13152"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/tags?post=13152"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}