Run semantic search, multi-modal retrieval, and knowledge-based Q&A directly in Lindorm—no data export needed. Results stay current and meet security, privacy, and regulatory requirements.
The AI engine provides these capabilities:
|
Vectorization for text or images Convert text or images from the wide table engine into vectors for vector search and clustering.  |
|
Semantic similarity calculation and reranking Calculate semantic similarity between a query and Lindorm Search results, then rerank by similarity score.  |
|
Retrieves semantically relevant content from Lindorm Search and the wide table engine, reranks results, and sends them with the user question to a large model to build a private Q&A service on your enterprise data.  |
Get started
Walk through vectorization, similarity calculation, reranking, and knowledge-based Q&A in one tutorial: Quickly build an intelligent search service based on the multi-modal capabilities of Lindorm.
Why choose the Lindorm AI engine
-
In-database data processing and AI inference
-
Ingestion, storage, computing, and inference all run inside Lindorm, keeping data in-database and meeting security, privacy, and compliance requirements.
-
No external data pipelines needed—this reduces complexity and maintenance costs. Inference runs on the latest data, improving freshness and accuracy.
-
-
Elastic heterogeneous computing
Inference nodes support CPU and GPU instance types. Use CPUs for standard tasks and GPUs for complex workloads. Inference nodes share storage with multi-modal engines, reducing data transfer overhead and enabling near-data optimization.
-
Seamless model platform integration
Import models from ModelScope and Hugging Face with one click, or upload custom models for specific business needs.
-
Native SQL interaction
Create AI models and run inference end-to-end using Lindorm SQL—no advanced programming required.
Supported models
Deploy open-source models from ModelScope and Hugging Face, or upload a custom model.
|
Model type |
Model list |
Model platform |
|
Vectorization (Embedding) |
|
Hugging Face |
|
jina-embeddings-v2-base-zh |
ModelScope |
|
|
Reranking (ReRank) |
|
Hugging Face |
|
Large language model (LLM) |
ChatGLM-6B |
Hugging Face |
|
ModelScope |
|
|
Text-to-image generation |
Stable Diffusion |
ModelScope |
|
Custom model |
N/A |
N/A |
Billing
The Lindorm AI engine requires AI foundation nodes or AI engine nodes. AI foundation nodes and model calls are free; AI engine nodes are paid.
The following table lists node requirements and billing per use case.
|
Use case |
Required nodes |
Billing |
|
AI inference services, including feature vector extraction, semantic search and reranking, Q&A, text generation, and text-to-image |
AI foundation node |
Free |
|
AI engine node |
Paid. Check the purchase page for pricing. |
AI foundation nodes manage and coordinate the AI engine. AI engine nodes deploy inference services.