{"id":30600604,"url":"https://github.com/microsoftcloudessentials-learninghub/azure-text-embedding-overview","last_synced_at":"2026-02-10T20:32:55.463Z","repository":{"id":311933116,"uuid":"1045664621","full_name":"MicrosoftCloudEssentials-LearningHub/Azure-Text-Embedding-Overview","owner":"MicrosoftCloudEssentials-LearningHub","description":"Optimize text embedding performance using Azure-native tools and models. For official guidance, support, or more detailed information, please refer to Microsoft's official documentation or contact Microsoft directly.","archived":false,"fork":false,"pushed_at":"2025-08-27T18:03:36.000Z","size":91,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-08-27T23:37:27.943Z","etag":null,"topics":["ai-foundry","ai-search","azure-openai","best-practices","marketplace","terraform-deployment","text-embedding"],"latest_commit_sha":null,"homepage":"","language":"HCL","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/MicrosoftCloudEssentials-LearningHub.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-08-27T14:26:22.000Z","updated_at":"2025-08-27T18:04:17.000Z","dependencies_parsed_at":"2025-08-27T23:37:29.255Z","dependency_job_id":"f2a6ff93-a21b-4517-ab2a-a05f339c6fb5","html_url":"https://github.com/MicrosoftCloudEssentials-LearningHub/Azure-Text-Embedding-Overview","commit_stats":null,"previous_names":["microsoftcloudessentials-learninghub/azure-text-embedding-bestpractices"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/MicrosoftCloudEssentials-LearningHub/Azure-Text-Embedding-Overview","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrosoftCloudEssentials-LearningHub%2FAzure-Text-Embedding-Overview","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrosoftCloudEssentials-LearningHub%2FAzure-Text-Embedding-Overview/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrosoftCloudEssentials-LearningHub%2FAzure-Text-Embedding-Overview/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrosoftCloudEssentials-LearningHub%2FAzure-Text-Embedding-Overview/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/MicrosoftCloudEssentials-LearningHub","download_url":"https://codeload.github.com/MicrosoftCloudEssentials-LearningHub/Azure-Text-Embedding-Overview/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrosoftCloudEssentials-LearningHub%2FAzure-Text-Embedding-Overview/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29314705,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-10T17:48:59.043Z","status":"ssl_error","status_checked_at":"2026-02-10T17:45:37.240Z","response_time":65,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-foundry","ai-search","azure-openai","best-practices","marketplace","terraform-deployment","text-embedding"],"created_at":"2025-08-29T23:38:03.021Z","updated_at":"2026-02-10T20:32:55.453Z","avatar_url":"https://github.com/MicrosoftCloudEssentials-LearningHub.png","language":"HCL","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Azure Text Embedding Model - Overview \n\nCosta Rica\n\n[![GitHub](https://badgen.net/badge/icon/github?icon=github\u0026label)](https://github.com)\n[![GitHub](https://img.shields.io/badge/--181717?logo=github\u0026logoColor=ffffff)](https://github.com/)\n[brown9804](https://github.com/brown9804)\n\nLast updated: 2025-08-27\n\n-----------------------------\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eList of References\u003c/b\u003e (Click to expand)\u003c/summary\u003e\n\n-  [voyage-3-large: the new state-of-the-art general-purpose embedding model](https://statics.teams.cdn.office.net/evergreen-assets/safelinks/2/atp-safelinks.html)\n-  [text-embedding-3-large](https://platform.openai.com/docs/models/text-embedding-3-large#:~:text=text-embedding-3-large%20is%20our%20most%20capable%20embedding,english%20and%20non-english%20tasks.)\n-  [Embedding Leaderboard](https://huggingface.co/spaces/mteb/leaderboard) - reports\n-  [The Legacy MTEB Leaderboard repository](https://github.com/embeddings-benchmark/leaderboard?tab=readme-ov-file) - explanation and relevant links \n-  [Massive Text Embedding Benchmark](https://github.com/embeddings-benchmark/mteb) - open source how it was created\n-  [Model leaderboards in Azure AI Foundry portal (preview)](https://learn.microsoft.com/en-us/azure/ai-foundry/concepts/model-benchmarks)\n-  [Find the best model by comparing model performance across various criteria](https://ai.azure.com/explore/models/leaderboard?tid=cc34547a-d117-4060-9c6e-1c8d622c8d02)\n-  [Chunk large documents for vector search solutions in Azure AI Search](https://learn.microsoft.com/en-us/azure/search/vector-search-how-to-chunk-documents)\n\n\u003c/details\u003e\n\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eTable of Contents\u003c/b\u003e (Click to expand)\u003c/summary\u003e\n\n- [Recommendations for Alternative \u0026 Upcoming Embedding Models in Azure](#recommendations-for-alternative--upcoming-embedding-models-in-azure)\n    - [Azure OpenAI](#azure-openai)\n    - [Azure AI Foundry](#azure-ai-foundry)\n    - [Azure Marketplace](#azure-marketplace)\n- [How Azure AI Search Can Help Optimize Embedding Performance](#how-azure-ai-search-can-help-optimize-embedding-performance)\n\n\u003c/details\u003e\n\n\n## Recommendations for Alternative \u0026 Upcoming Embedding Models in Azure\n\n\u003e Current Options: \n\n| **Model** | **Platform** | **Strengths** | **Limitations** | **Best Use Cases** |\n|-----------|--------------|---------------|------------------|---------------------|\n| `voyage-3` | Azure Marketplace | Balanced performance, tuned for enterprise domains, cost-efficient | No native integration with Azure AI Foundry, limited multilingual support | Internal search, domain-specific retrieval, enterprise RAG |\n| `voyage-3-lite` | Azure Marketplace | Lightweight, fast, low latency and compute cost | Slightly reduced semantic precision, less robust for long-form text | Mobile apps, real-time classification, edge deployments |\n| `voyage-finance-2` | Azure Marketplace | Specialized for financial language and jargon, tuned for structured documents | Poor generalization outside finance, narrow scope | Financial document analysis, compliance, financial RAG |\n| `voyage-law-2` | Azure Marketplace | Legal-specific, optimized for contracts and statutes | Limited cross-domain utility, narrow vocabulary | Legal search, contract parsing, regulatory compliance |\n| `voyage-multilingual-2` | Azure Marketplace | Multilingual support, tuned for global corpora | Slightly lower precision in English, not ideal for domain-specific tasks | Multilingual semantic search, global content indexing |\n| `text-embedding-ada-002` | Azure OpenAI / AI Foundry | Legacy model, fast and inexpensive, widely supported | Lower semantic quality, outdated compared to newer models | Lightweight search, prototyping, low-cost RAG |\n| `text-embedding-3-small` | Azure OpenAI / AI Foundry | Fast, cost-effective, configurable, good for short texts | Lower accuracy on nuanced or complex queries | Chat summarization, real-time applications, indexing |\n| `text-embedding-3-large` | Azure OpenAI / AI Foundry | High semantic fidelity, multilingual, robust across domains | Higher latency and cost, batch quirks | Semantic search, multilingual corpora, recommendation systems |\n| `embed-v-4-0` | Azure AI Foundry | High-performance, optimized for retrieval and RAG, scalable | Newer model with limited public benchmarks | Advanced RAG pipelines, enterprise search, hybrid retrieval |\n| `Cohere-embed-v3-multilingual` | Azure AI Foundry | Strong multilingual alignment, semantic robustness | Slightly slower, less tuned for English-only tasks | Multilingual indexing, global search, translation-aware retrieval |\n| `Cohere-embed-v3-english` | Azure AI Foundry | High precision for English, optimized for semantic tasks | Not suitable for multilingual content | English-centric semantic search, document clustering |\n\n\u003e [!NOTE]\n\u003e `Upcoming Models to Watch:`\n- **Voyage-3-Large**: Expected to outperform OpenAI v3-large with flexible dimensions and quantization. [voyage-3-large: the new state-of-the-art general-purpose embedding model](https://statics.teams.cdn.office.net/evergreen-assets/safelinks/2/atp-safelinks.html)\n\n### Azure OpenAI\n\n\u003e - `text-embedding-ada-002`\n\u003e - `text-embedding-3-large`\n\u003e - `text-embedding-3-small`\n\n1. Go to your Azure OpenAI Platform:\n\n     \u003cimg width=\"1916\" height=\"840\" alt=\"image\" src=\"https://github.com/user-attachments/assets/424f2d85-37e4-4e74-b303-2e8c5a23a864\" /\u003e\n\n2. Under `Model catalog`, filter by `Inference task` \u003e `Embeddings`:\n\n     \u003cimg width=\"1905\" height=\"848\" alt=\"image\" src=\"https://github.com/user-attachments/assets/8747ad65-bf2b-4fdd-bdb8-6ad505b4ebdf\" /\u003e\n\n     \u003cimg width=\"1893\" height=\"695\" alt=\"image\" src=\"https://github.com/user-attachments/assets/8d822c1d-04fd-49ba-a514-1a3f74e24cdc\" /\u003e\n\n### Azure AI Foundry\n\n\u003e - `embed-v-4-0`\n\u003e - `Cohere-embed-v3-multilingual`\n\u003e - `Cohere-embed-v3-english`\n\u003e - `text-embedding-ada-002`\n\u003e - `text-embedding-3-large`\n\u003e - `text-embedding-3-small`\n\n1. Go to your Azure AI Foundry Platform:\n\n    \u003cimg width=\"1897\" height=\"791\" alt=\"image\" src=\"https://github.com/user-attachments/assets/f56fc29b-2276-4998-9f0c-226e4329ca78\" /\u003e\n\n2. Under `Model catalog`, filter by `Inference task` \u003e `Embeddings`:\n\n    \u003cimg width=\"1907\" height=\"847\" alt=\"image\" src=\"https://github.com/user-attachments/assets/45b1676c-8c5b-4084-be19-4eeb56e8f961\" /\u003e\n\n### Azure Marketplace\n\n\u003e [Azure Marketplace](https://azuremarketplace.microsoft.com/en-us/marketplace/apps?search=embedding\u0026page=1):\n\u003e - `voyage-3 Embedding Model`\n\u003e - `voyage-3-lite Embedding Model`\n\u003e - `voyage-finance-2 Embedding Model`\n\u003e - `voyage-law-2 Embedding Model`\n\u003e - `voyage-multilingual-2 Embedding Model`\n\n\u003cimg width=\"1900\" height=\"837\" alt=\"image\" src=\"https://github.com/user-attachments/assets/0eac2f90-bf75-45cf-aa06-5e4420dc8be5\" /\u003e\n\n## How Azure AI Search Can Help Optimize Embedding Performance\n\n\u003e Azure AI Search doesn’t generate embeddings from third-party models, but it **amplifies their value** through advanced indexing, retrieval, and hybrid search capabilities.\n\n\u003e [!TIP]\n\u003e Use Azure AI Foundry to experiment with multiple embedding models and benchmark their performance before committing to production.\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eVector Search Capabilities\u003c/b\u003e (Click to expand)\u003c/summary\u003e\n\n\u003e Azure AI Search supports robust vector search features that allow you to fully leverage embeddings generated externally (e.g., from OpenAI, Voyage, Cohere):\n\n- Accepts **custom embeddings** from any model, making it agnostic and flexible across providers.\n- Supports **semantic similarity search** using metrics like cosine similarity and dot product, enabling nuanced matching beyond keyword overlap.\n- Handles **large-scale indexing** and **low-latency retrieval**, ideal for enterprise-grade applications with millions of documents.\n- Integrates with **hybrid search pipelines**, combining vector and keyword search to improve both precision and recall.\n\n\u003e Example Use Cases:\n\n- Intelligent document retrieval for legal or financial archives.\n- FAQ matching and chatbot grounding using semantic similarity.\n- Product recommendation systems based on user intent embeddings.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003ePerformance Optimization Tips\u003c/b\u003e (Click to expand)\u003c/summary\u003e\n\n\u003e To maximize efficiency and relevance in embedding-based search, consider these strategies:\n\n1. **Choose embedding size wisely**: Smaller embeddings (e.g., 512–1024) reduce latency and storage costs, ideal for mobile or real-time apps.\n2. **Batch embedding generation**: Pre-process documents in bulk to reduce API calls and improve throughput.\n3. **Use domain-specific models**: Models like `voyage-finance-2` or `voyage-law-2` yield better semantic relevance in specialized contexts.\n4. **Monitor vector DB costs**: Larger embeddings increase storage and query costs, balance precision with efficiency.\n5. **Leverage hybrid search**: Combine keyword and vector search to handle both exact and fuzzy matches, especially in noisy datasets.\n6. **Normalize and deduplicate embeddings**: Ensure consistent vector quality and avoid redundant indexing.\n\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cb\u003eIntegration Strategy\u003c/b\u003e (Click to expand)\u003c/summary\u003e\n\n\u003e Azure AI Search is designed to integrate seamlessly with external embedding pipelines. Here's how to build a modular and scalable setup:\n\n- **Generate embeddings externally** using models from Azure OpenAI (`text-embedding-3-large`, `text-embedding-3-small`), Voyage AI (`voyage-3`, `voyage-multilingual-2`), or Cohere (`embed-v3` series).\n- **Store embeddings** in Azure AI Search vector fields, which are optimized for fast similarity search and scalable indexing.\n- **Query using embedded vectors** to perform semantic matching, enabling intelligent document retrieval, contextual search, and RAG workflows.\n- **Automate updates**: Use Azure Functions or Logic Apps to refresh embeddings when documents change, keeping your index up-to-date.\n\n\u003e Example Workflow:\n\n1. Use `text-embedding-3-large` to embed support tickets.\n2. Store vectors in Azure AI Search.\n3. Query with user questions to retrieve semantically similar tickets.\n4. Combine with keyword filters for precision.\n\n\u003c/details\u003e\n\n\n\u003c!-- START BADGE --\u003e\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"https://img.shields.io/badge/Total%20views-1304-limegreen\" alt=\"Total views\"\u003e\n  \u003cp\u003eRefresh Date: 2025-08-27\u003c/p\u003e\n\u003c/div\u003e\n\u003c!-- END BADGE --\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmicrosoftcloudessentials-learninghub%2Fazure-text-embedding-overview","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmicrosoftcloudessentials-learninghub%2Fazure-text-embedding-overview","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmicrosoftcloudessentials-learninghub%2Fazure-text-embedding-overview/lists"}