使用 Jina Reader 新的 Grounding API 進行事實查核

curl -X POST https://g.jina.ai \
     -H "Content-Type: application/json" \
     -H "Authorization: Bearer $YOUR_JINA_TOKEN" \
     -d '{
           "statement":"the last model released by Jina AI is jina-embeddings-v3"
         }'
Copy

{
  "code": 200,
  "status": 20000,
  "data": {
    "factuality": 0.95,
    "result": true,
    "reason": "The majority of the references explicitly support the statement that the last model released by Jina AI is jina-embeddings-v3. Multiple sources, such as the arXiv paper, Jina AI's news, and various model documentation pages, confirm this assertion. Although there are a few references to the jina-embeddings-v2 model, they do not provide evidence contradicting the release of a subsequent version (jina-embeddings-v3). Therefore, the statement that 'the last model released by Jina AI is jina-embeddings-v3' is well-supported by the provided documentation.",
    "references": [
      {
        "url": "https://arxiv.org/abs/2409.10173",
        "keyQuote": "arXiv September 18, 2024 jina-embeddings-v3: Multilingual Embeddings With Task LoRA",
        "isSupportive": true
      },
      {
        "url": "https://arxiv.org/abs/2409.10173",
        "keyQuote": "We introduce jina-embeddings-v3, a novel text embedding model with 570 million parameters, achieves state-of-the-art performance on multilingual data and long-context retrieval tasks, supporting context lengths of up to 8192 tokens.",
        "isSupportive": true
      },
      {
        "url": "https://azuremarketplace.microsoft.com/en-us/marketplace/apps/jinaai.jina-embeddings-v3?tab=Overview",
        "keyQuote": "jina-embeddings-v3 is a multilingual multi-task text embedding model designed for a variety of NLP applications.",
        "isSupportive": true
      },
      {
        "url": "https://docs.pinecone.io/models/jina-embeddings-v3",
        "keyQuote": "Jina Embeddings v3 is the latest iteration in the Jina AI's text embedding model series, building upon Jina Embedding v2.",
        "isSupportive": true
      },
      {
        "url": "https://haystack.deepset.ai/integrations/jina",
        "keyQuote": "Recommended Model: jina-embeddings-v3 : We recommend jina-embeddings-v3 as the latest and most performant embedding model from Jina AI.",
        "isSupportive": true
      },
      {
        "url": "https://huggingface.co/jinaai/jina-embeddings-v2-base-en",
        "keyQuote": "The embedding model was trained using 512 sequence length, but extrapolates to 8k sequence length (or even longer) thanks to ALiBi.",
        "isSupportive": false
      },
      {
        "url": "https://huggingface.co/jinaai/jina-embeddings-v2-base-en",
        "keyQuote": "With a standard size of 137 million parameters, the model enables fast inference while delivering better performance than our small model.",
        "isSupportive": false
      },
      {
        "url": "https://huggingface.co/jinaai/jina-embeddings-v2-base-en",
        "keyQuote": "We offer an `encode` function to deal with this.",
        "isSupportive": false
      },
      {
        "url": "https://huggingface.co/jinaai/jina-embeddings-v3",
        "keyQuote": "jinaai/jina-embeddings-v3 Feature Extraction • Updated 3 days ago • 278k • 375",
        "isSupportive": true
      },
      {
        "url": "https://huggingface.co/jinaai/jina-embeddings-v3",
        "keyQuote": "the latest version (3.1.0) of [SentenceTransformers] also supports jina-embeddings-v3",
        "isSupportive": true
      },
      {
        "url": "https://huggingface.co/jinaai/jina-embeddings-v3",
        "keyQuote": "jina-embeddings-v3: Multilingual Embeddings With Task LoRA",
        "isSupportive": true
      },
      {
        "url": "https://jina.ai/embeddings/",
        "keyQuote": "v3: Frontier Multilingual Embeddings is a frontier multilingual text embedding model with 570M parameters and 8192 token-length, outperforming the latest proprietary embeddings from OpenAI and Cohere on MTEB.",
        "isSupportive": true
      },
      {
        "url": "https://jina.ai/news/jina-embeddings-v3-a-frontier-multilingual-embedding-model",
        "keyQuote": "Jina Embeddings v3: A Frontier Multilingual Embedding Model jina-embeddings-v3 is a frontier multilingual text embedding model with 570M parameters and 8192 token-length, outperforming the latest proprietary embeddings from OpenAI and Cohere on MTEB.",
        "isSupportive": true
      },
      {
        "url": "https://jina.ai/news/jina-embeddings-v3-a-frontier-multilingual-embedding-model/",
        "keyQuote": "As of its release on September 18, 2024, jina-embeddings-v3 is the best multilingual model ...",
        "isSupportive": true
      }
    ],
    "usage": {
      "tokens": 112073
    }
  }
}
Copy