feat(provider): add DashScope text-embedding-v4 embedding provider

astrbot/core/config/default.py

@@ -1825,6 +1825,20 @@
                         "timeout": 60,
                         "proxy": "",
                     },
+                    "阿里云百炼 Embedding": {
+                        "id": "dashscope_embedding",
+                        "type": "dashscope_embedding",
+                        "provider": "dashscope",
+                        "provider_type": "embedding",
+                        "hint": "provider_group.provider.dashscope_embedding.hint",
+                        "enable": True,
+                        "embedding_api_key": "",
+                        "embedding_api_base": "https://dashscope.aliyuncs.com/compatible-mode/v1",
+                        "embedding_model": "text-embedding-v4",
+                        "embedding_dimensions": 1024,
+                        "timeout": 20,
+                        "proxy": "",
+                    },
                     "vLLM Rerank": {
                         "id": "vllm_rerank",
                         "type": "vllm_rerank",

astrbot/core/provider/manager.py

@@ -477,6 +477,10 @@ def dynamic_import_provider(self, type: str) -> None:
                 from .sources.ollama_embedding_source import (
                     OllamaEmbeddingProvider as OllamaEmbeddingProvider,
                 )
+            case "dashscope_embedding":
+                from .sources.dashscope_embedding_source import (
+                    DashscopeEmbeddingProvider as DashscopeEmbeddingProvider,
+                )
             case "vllm_rerank":
                 from .sources.vllm_rerank_source import (
                     VLLMRerankProvider as VLLMRerankProvider,

astrbot/core/provider/sources/dashscope_embedding_source.py

@@ -0,0 +1,88 @@
+import httpx
+from openai import AsyncOpenAI
+
+from astrbot import logger
+
+from ..entities import ProviderType
+from ..provider import EmbeddingProvider
+from ..register import register_provider_adapter
+
+
+@register_provider_adapter(
+    "dashscope_embedding",
+    "DashScope text-embedding-v4 Embedding Provider",
+    provider_type=ProviderType.EMBEDDING,
+)
+class DashscopeEmbeddingProvider(EmbeddingProvider):
+    def __init__(self, provider_config: dict, provider_settings: dict) -> None:
+        super().__init__(provider_config, provider_settings)
+        proxy = provider_config.get("proxy", "")
+        provider_id = provider_config.get("id", "unknown_id")
+        http_client = None
+        if proxy:
+            logger.info(f"[DashScope Embedding] {provider_id} Using proxy: {proxy}")
+            http_client = httpx.AsyncClient(proxy=proxy)
+        api_base = (
+            provider_config.get(
+                "embedding_api_base",
+                "https://dashscope.aliyuncs.com/compatible-mode/v1",
+            )
+            .strip()
+            .removesuffix("/")
+            .removesuffix("/embeddings")
+        )
+        if api_base and not api_base.endswith("/v1") and not api_base.endswith("/v4"):
+            api_base = api_base + "/v1"
+        logger.info(f"[DashScope Embedding] {provider_id} Using API Base: {api_base}")
+        self.client = AsyncOpenAI(
+            api_key=provider_config.get("embedding_api_key"),
+            base_url=api_base,
+            timeout=int(provider_config.get("timeout", 20)),
+            http_client=http_client,
+        )
+        self.model = provider_config.get("embedding_model", "text-embedding-v4")
+
+    async def get_embedding(self, text: str) -> list[float]:
+        kwargs = self._embedding_kwargs()
+        embedding = await self.client.embeddings.create(
+            input=text,
+            model=self.model,
+            **kwargs,
+        )
+        if not embedding.data:
+            raise Exception("DashScope API returned no embedding data.")
+        return embedding.data[0].embedding
+
+    async def get_embeddings(self, text: list[str]) -> list[list[float]]:
+        kwargs = self._embedding_kwargs()
+        embeddings = await self.client.embeddings.create(
+            input=text,
+            model=self.model,
+            **kwargs,
+        )
+        return [item.embedding for item in embeddings.data]
+
+    def _embedding_kwargs(self) -> dict:
+        kwargs = {}
+        if "embedding_dimensions" in self.provider_config:
+            try:
+                kwargs["dimensions"] = int(self.provider_config["embedding_dimensions"])
+            except (ValueError, TypeError):
+                logger.warning(
+                    f"embedding_dimensions in embedding configs is not a valid integer: '{self.provider_config['embedding_dimensions']}', ignored."
+                )
+        return kwargs
+
+    def get_dim(self) -> int:
+        if "embedding_dimensions" in self.provider_config:
+            try:
+                return int(self.provider_config["embedding_dimensions"])
+            except (ValueError, TypeError):
+                logger.warning(
+                    f"embedding_dimensions in embedding configs is not a valid integer: '{self.provider_config['embedding_dimensions']}', ignored."
+                )
+        return 1024
+
+    async def terminate(self):
+        if self.client:
+            await self.client.close()

dashboard/src/i18n/locales/en-US/features/config-metadata.json

@@ -1321,6 +1321,9 @@
       "gemini_embedding": {
         "hint": "Gemini Embedding does not require manually adding /v1beta."
       },
+      "dashscope_embedding": {
+        "hint": "text-embedding-v4 supports multiple dimensions: 64, 128, 256, 512, 768, 1024 (default), 1536, 2048. API base defaults to DashScope compatible mode."
+      },
       "volcengine_cluster": {
         "description": "Volcengine cluster",
         "hint": "For voice cloning models, choose volcano_icl or volcano_icl_concurr; default is volcano_tts."

dashboard/src/i18n/locales/ru-RU/features/config-metadata.json

@@ -1318,6 +1318,9 @@
             "gemini_embedding": {
                 "hint": "Gemini Embedding не требует ручного добавления /v1beta."
             },
+            "dashscope_embedding": {
+                "hint": "text-embedding-v4 поддерживает несколько размеров: 64, 128, 256, 512, 768, 1024 (по умолчанию), 1536, 2048. API base по умолчанию использует совместимый режим DashScope."
+            },
             "volcengine_cluster": {
                 "description": "Кластер Volcengine",
                 "hint": "Для моделей клонирования голоса выберите volcano_icl или volcano_icl_concurr; по умолчанию volcano_tts."

dashboard/src/i18n/locales/zh-CN/features/config-metadata.json

@@ -1323,6 +1323,9 @@
       "gemini_embedding": {
         "hint": "Gemini Embedding 无需手动添加 /v1beta。"
       },
+      "dashscope_embedding": {
+        "hint": "text-embedding-v4 支持多种维度: 64、128、256、512、768、1024（默认）、1536、2048。API Base 默认使用 DashScope 兼容模式。"
+      },
       "volcengine_cluster": {
         "description": "火山引擎集群",
         "hint": "若使用语音复刻大模型，可选volcano_icl或volcano_icl_concurr，默认使用volcano_tts"