| {"access": "api", "api_key_env": "OPENAI_API_KEY", "dimensions": 3072, "display_name": "OpenAI text-embedding-3-large", "id": "openai-text-embedding-3-large", "max_text_length": 8191, "modalities": ["text"], "notes": null, "priority": 10, "provider": "openai", "provider_kwargs": {"model": "text-embedding-3-large"}, "publish": true, "source": "https://platform.openai.com/docs/guides/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "baseline", "text", "mrl"]} |
| {"access": "api", "api_key_env": "OPENAI_API_KEY", "dimensions": 1536, "display_name": "OpenAI text-embedding-3-small", "id": "openai-text-embedding-3-small", "max_text_length": 8191, "modalities": ["text"], "notes": null, "priority": 20, "provider": "openai", "provider_kwargs": {"model": "text-embedding-3-small"}, "publish": true, "source": "https://platform.openai.com/docs/guides/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "low-cost", "text", "mrl"]} |
| {"access": "api", "api_key_env": "VOYAGE_API_KEY", "dimensions": 1024, "display_name": "Voyage 4 Large", "id": "voyage-4-large", "max_text_length": 32000, "modalities": ["text"], "notes": "Supports 256, 512, 1024, and 2048 output dimensions.", "priority": 11, "provider": "voyage", "provider_kwargs": {"model": "voyage-4-large"}, "publish": true, "source": "https://docs.voyageai.com/docs/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multilingual", "text", "mrl", "voyage"]} |
| {"access": "api", "api_key_env": "VOYAGE_API_KEY", "dimensions": 1024, "display_name": "Voyage 4", "id": "voyage-4", "max_text_length": 32000, "modalities": ["text"], "notes": "Supports 256, 512, 1024, and 2048 output dimensions.", "priority": 12, "provider": "voyage", "provider_kwargs": {"model": "voyage-4"}, "publish": true, "source": "https://docs.voyageai.com/docs/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multilingual", "text", "mrl", "voyage"]} |
| {"access": "api", "api_key_env": "VOYAGE_API_KEY", "dimensions": 1024, "display_name": "Voyage 4 Lite", "id": "voyage-4-lite", "max_text_length": 32000, "modalities": ["text"], "notes": "Supports 256, 512, 1024, and 2048 output dimensions.", "priority": 13, "provider": "voyage", "provider_kwargs": {"model": "voyage-4-lite"}, "publish": true, "source": "https://docs.voyageai.com/docs/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "low-cost", "multilingual", "text", "mrl", "voyage"]} |
| {"access": "api", "api_key_env": "VOYAGE_API_KEY", "dimensions": 1024, "display_name": "Voyage Multimodal 3.5", "id": "voyage-multimodal-3.5", "max_text_length": 32000, "modalities": ["text", "image", "document"], "notes": "Current provider path covers text, image, and document inputs; video support needs provider work before benchmark inclusion.", "priority": 18, "provider": "voyage", "provider_kwargs": {"model": "voyage-multimodal-3.5"}, "publish": true, "source": "https://docs.voyageai.com/docs/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multimodal", "text", "image", "document", "mrl", "voyage"]} |
| {"access": "api", "api_key_env": "GEMINI_API_KEY", "dimensions": 3072, "display_name": "Gemini Embedding 2", "id": "gemini-embedding-2", "max_text_length": 8192, "modalities": ["text", "image", "audio", "video", "document"], "notes": "Embeddings 2 retrieval specialization uses prompt instructions rather than the legacy task_type field.", "priority": 14, "provider": "gemini", "provider_kwargs": {"model": "gemini-embedding-2"}, "publish": true, "source": "https://ai.google.dev/gemini-api/docs/embeddings", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multimodal", "text", "image", "audio", "video", "document", "mrl", "gemini"]} |
| {"access": "api", "api_key_env": "JINA_API_KEY", "dimensions": 1024, "display_name": "Jina Embeddings v5 Text Small", "id": "jina-embeddings-v5-text-small", "max_text_length": 32768, "modalities": ["text"], "notes": "v5 text family supports task-specific LoRA adapters and Matryoshka dimensions.", "priority": 15, "provider": "jina", "provider_kwargs": {"model": "jina-embeddings-v5-text-small"}, "publish": true, "source": "https://jina.ai/models/jina-embeddings-v5-text-small/", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multilingual", "text", "mrl", "jina"]} |
| {"access": "api", "api_key_env": "JINA_API_KEY", "dimensions": 768, "display_name": "Jina Embeddings v5 Text Nano", "id": "jina-embeddings-v5-text-nano", "max_text_length": 8192, "modalities": ["text"], "notes": "Smallest v5 text API candidate for cheap smoke runs.", "priority": 16, "provider": "jina", "provider_kwargs": {"model": "jina-embeddings-v5-text-nano"}, "publish": true, "source": "https://jina.ai/models/jina-embeddings-v5-text-nano/", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "low-cost", "multilingual", "text", "mrl", "jina"]} |
| {"access": "api", "api_key_env": "JINA_API_KEY", "dimensions": 2048, "display_name": "Jina Embeddings v4", "id": "jina-embeddings-v4", "max_text_length": 32768, "modalities": ["text", "image", "document"], "notes": "Supports task-specific LoRA adapters and multimodal embeddings.", "priority": 19, "provider": "jina", "provider_kwargs": {"model": "jina-embeddings-v4"}, "publish": true, "source": "https://jina.ai/models/jina-embeddings-v4/", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multimodal", "multilingual", "text", "image", "document", "mrl", "jina"]} |
| {"access": "api", "api_key_env": "COHERE_API_KEY", "dimensions": 1536, "display_name": "Cohere Embed v4.0", "id": "cohere-embed-v4", "max_text_length": 128000, "modalities": ["text", "image", "document"], "notes": "Local loop snapshot reports recent 403 access errors; keep out of default smoke runs until rechecked.", "priority": 17, "provider": "cohere", "provider_kwargs": {"model": "embed-v4.0"}, "publish": true, "source": "https://docs.cohere.com/docs/cohere-embed", "status": "access_recheck", "supports_mrl": true, "tags": ["api", "modern", "multimodal", "multilingual", "text", "image", "document", "mrl", "cohere"]} |
| {"access": "api", "api_key_env": "COHERE_API_KEY", "dimensions": 1024, "display_name": "Cohere Embed Multilingual v3.0", "id": "cohere-embed-multilingual-v3", "max_text_length": 512, "modalities": ["text"], "notes": "Text-only fallback candidate; keep in access recheck until Cohere account/API path is verified.", "priority": 21, "provider": "cohere", "provider_kwargs": {"model": "embed-multilingual-v3.0"}, "publish": true, "source": "https://docs.cohere.com/docs/cohere-embed", "status": "access_recheck", "supports_mrl": false, "tags": ["api", "multilingual", "text", "cohere"]} |
| {"access": "api", "api_key_env": "DASHSCOPE_API_KEY", "dimensions": 1024, "display_name": "DashScope text-embedding-v4", "id": "dashscope-text-embedding-v4", "max_text_length": 8192, "modalities": ["text"], "notes": "Qwen3-Embedding API model; registry uses the provider's 1024-dim default smoke setting.", "priority": 22, "provider": "dashscope", "provider_kwargs": {"model": "text-embedding-v4"}, "publish": true, "source": "https://www.alibabacloud.com/help/en/model-studio/embedding", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multilingual", "text", "mrl", "dashscope", "qwen"]} |
| {"access": "api", "api_key_env": "DASHSCOPE_API_KEY", "dimensions": 1024, "display_name": "DashScope Qwen3-VL-Embedding", "id": "dashscope-qwen3-vl-embedding", "max_text_length": 8192, "modalities": ["text", "image", "video"], "notes": "Provider routes single text, image, or video items through MultiModalEmbedding; fused multi-item inputs are not covered by this smoke path.", "priority": 23, "provider": "dashscope", "provider_kwargs": {"model": "qwen3-vl-embedding"}, "publish": true, "source": "https://www.alibabacloud.com/help/en/model-studio/embedding", "status": "active", "supports_mrl": true, "tags": ["api", "modern", "multimodal", "multilingual", "text", "image", "video", "mrl", "dashscope", "qwen"]} |
| {"access": "api", "api_key_env": "ARK_API_KEY", "dimensions": 2048, "display_name": "ARK Doubao Embedding Large Text 250515", "id": "ark-doubao-embedding-large-text-250515", "max_text_length": 8192, "modalities": ["text"], "notes": "Text embedding model available through the Volcengine ARK OpenAI-compatible endpoint.", "priority": 24, "provider": "ark", "provider_kwargs": {"model": "doubao-embedding-large-text-250515"}, "publish": true, "source": "https://www.volcengine.com/docs/82379/1330310", "status": "active", "supports_mrl": false, "tags": ["api", "modern", "multilingual", "text", "ark", "doubao", "volcengine"]} |
| {"access": "api", "api_key_env": "ARK_API_KEY", "dimensions": 1024, "display_name": "Volcengine Doubao Embedding Vision 251215", "id": "volcengine-doubao-embedding-vision-251215", "max_text_length": 32768, "modalities": ["text", "image", "video"], "notes": "Registered for follow-up only; endpoint/deployment compatibility and MRL dimension behavior still need validation.", "priority": 25, "provider": "volcengine", "provider_kwargs": {"model": "doubao-embedding-vision-251215"}, "publish": true, "source": "https://www.volcengine.com/docs/82379/1409291", "status": "access_recheck", "supports_mrl": false, "tags": ["api", "modern", "multimodal", "multilingual", "text", "image", "video", "volcengine", "doubao"]} |
| {"access": "open-weight", "api_key_env": null, "dimensions": 1024, "display_name": "BAAI/bge-m3", "id": "bge-m3-local", "max_text_length": 8192, "modalities": ["text"], "notes": null, "priority": 30, "provider": "sentence_transformers", "provider_kwargs": {"model": "BAAI/bge-m3"}, "publish": true, "source": "https://huggingface.co/BAAI/bge-m3", "status": "active", "supports_mrl": true, "tags": ["local", "open-weight", "multilingual", "text", "mrl"]} |
| {"access": "open-weight", "api_key_env": null, "dimensions": 512, "display_name": "CLIP ViT-B/32", "id": "clip-vit-b-32-local", "max_text_length": 77, "modalities": ["text", "image"], "notes": null, "priority": 60, "provider": "sentence_transformers", "provider_kwargs": {"model": "clip-ViT-B-32"}, "publish": true, "source": "https://huggingface.co/sentence-transformers/clip-ViT-B-32", "status": "active", "supports_mrl": false, "tags": ["local", "open-weight", "multimodal", "clip"]} |
|
|
