# FAQ

Get answers to common questions about Spice Cloud, Spice.ai OSS, and Spice.ai Enterprise, including pricing, support, SQL, and CDC.

## What's the difference between Spice.ai OSS, Cloud, and Enterprise?

- **Spice.ai OSS**: the open-source Spice runtime. Self-hosted, free, Apache 2.0.
- **Spice Cloud**: a managed, multi-tenant hosted service running Spice as a platform with additional building blocks (cloud data warehouse, model training/inference, AI gateway).
- **Spice.ai Enterprise**: a self-hosted enterprise distribution of Spice with advanced features (HA, RBAC, SSO, governance, premium connectors), enterprise support, and an SLA.

See [Distributions](https://spice.ai/docs/enterprise/getting-started/distributions) for a detailed comparison.

## What's the difference between the Spice Cloud Platform and Spice.ai OSS?

[**Spice.ai OSS**](https://github.com/spiceai/spiceai) is an open-source project created by the Spice AI team that provides a unified SQL query interface to locally materialize, accelerate, and query data tables sourced from any database, data warehouse, or data lake.

![A three-layer architecture diagram. Data and AI applications, services, dashboards, and BI tools connect to the runtime over HTTP, Arrow Flight (gRPC), JDBC, ODBC, or ADBC. The runtime holds a federated SQL query engine (Apache DataFusion) and data materialization (DuckDB, SQLite). It also holds ML inference and vector search (PostgreSQL, pgvector) and an AI gateway (OpenAI, Azure). Monitoring, observability, control, and policy cover all four parts. The bottom layer holds structured data (Snowflake, Databricks), unstructured data (SharePoint, FTP), and models (ONNX, Hugging Face).](https://spice.ai/website-assets/docs/image-41.png)

The Spice.ai OSS runtime

**Spice Cloud** is a data and AI application platform that provides a set of building-blocks to create AI and agentic applications. Building blocks include a cloud-data-warehouse, ML model training and inference, and a cloud-scale, managed Spice.ai OSS cloud-hosted service.

![A platform diagram for data and time-series AI applications. The top row shows SDKs for Node.js, Python, Go, and Rust. It also shows libraries such as pandas, PyTorch, TensorFlow, and Apache Arrow. The middle panel shows five building blocks: SQL Query, ML Models, Functions, Hosted Nodes, and Web3 Data. A Spice.ai OSS tag sits across the SQL Query and ML Models blocks. The bottom row lists a petabyte-scale data platform, AI/ML training and inferencing, blockchain nodes with indexing and enrichment, and bare-metal cloud.](https://spice.ai/website-assets/docs/ogp-1.png)

The Spice Cloud Platform

## How much does Spice Cloud cost?

It's free to [get an API key](https://spice.ai/) to use the [Community Edition](https://spice.ai/docs/cloud/pricing/community).

Customers who need resource limits, service-level guarantees, or priority support we offer [high-value paid tiers](https://spice.ai/docs/cloud/pricing/plans) based on usage.

## What level of support do you offer?

We offer enterprise-grade support with an SLA [for Enterprise Plans](https://spice.ai/docs/cloud/pricing/plans).

For standard plans we offer [best-effort community support](https://spice.ai/slack) in the Spice AI Community Slack.

## What's your approach to security and compliance?

See [Security](https://spice.ai/security). Spice Cloud is SOC 2 Type II compliant.

## What SQL query engine/dialect do you support?

Spice.ai OSS is built on [Apache DataFusion](https://datafusion.apache.org/) as its primary query execution engine, providing vectorized, multi-threaded query processing. It uses the PostgreSQL SQL dialect. Spice also supports [DuckDB](https://duckdb.org/), SQLite, and PostgreSQL as acceleration engines at the dataset level.

## What AI capabilities does Spice provide?

Spice provides unified APIs for data and AI workflows, including model inference, embeddings, and an [AI gateway](https://spice.ai/docs/features/ai-gateway) supporting OpenAI, Anthropic, Amazon Bedrock, and xAI. Spice also includes advanced tools such as [vector and hybrid search](https://spice.ai/docs/features/search-and-retrieval), text-to-SQL, and data sampling.

## What AI model providers does Spice support?

Spice supports local model serving (e.g. Llama) and gateways to hosted AI platforms including OpenAI, Anthropic, xAI, and Amazon Bedrock. See [Model Providers](https://spice.ai/docs/building-blocks/model-providers) for details.

## Can Spice handle federated queries?

Yes. Spice natively supports [federated SQL queries](https://spice.ai/docs/features/federated-sql-query) across disparate data sources with advanced query push-down capabilities, executing portions of queries directly on source databases to reduce data transfer and improve performance.

## Can Spice integrate with existing BI tools?

Yes. Spice integrates with BI tools through standard SQL interfaces (ODBC, JDBC, ADBC, Arrow Flight SQL), enabling accelerated, real-time analytics for dashboards and reporting.

## Does Spice support Change Data Capture (CDC)?

Yes. Spice supports streaming ingestion from several sources:

- **Native PostgreSQL logical replication** (recommended for PostgreSQL sources). Spice connects directly to the source using Postgres' `wal_level=logical` and streams `INSERT`/`UPDATE`/`DELETE` events into the accelerator. See [PostgreSQL Logical Replication](https://spiceai.org/docs/features/cdc/postgres-replication) in the OSS documentation.
- **[MySQL binlog replication](https://spice.ai/docs/building-blocks/data-connectors/mysql)** for MySQL sources. Spice loads an initial snapshot of the table, then applies inserts, updates, and deletes from the MySQL binary log with `refresh_mode: changes`.
- **[MongoDB change streams](https://spice.ai/docs/building-blocks/data-connectors/mongodb)** for MongoDB collections, with `refresh_mode: changes`.
- **[DynamoDB Streams](https://spice.ai/docs/building-blocks/data-connectors/dynamodb)** for Amazon DynamoDB sources. Spice consumes the table's change stream and applies `INSERT`/`UPDATE`/`DELETE` events to the accelerator with `refresh_mode: changes`.
- **[Apache Kafka](https://spice.ai/docs/building-blocks/data-connectors/kafka)** for event-streaming topics. Spice consumes records directly with `refresh_mode: append` for real-time, append-only acceleration.
- **[Debezium](https://spice.ai/docs/building-blocks/data-connectors/debezium)** (over Kafka) for sources where Debezium is already deployed, or for databases without a native Spice CDC path (SQL Server, etc.).

## How do I keep an accelerated dataset incrementally up-to-date?

For sources with a monotonically-increasing version column (e.g. `updated_at`), Spice incrementally ingests new and modified records using `time_column` + `refresh_mode: append`, with `refresh_append_overlap` to tolerate clock skew and `retention_period` to evict old or soft-deleted records. See [Incremental Ingestion](https://spice.ai/docs/features/data-acceleration#incremental-ingestion) for configuration details and examples.

## Does Spice support schema evolution?

Spice infers the schema for datasets and views at startup. How the runtime responds to a source schema change while running (e.g. columns added, removed, or retyped) is controlled per dataset by the `on_schema_change` setting.

By default (`on_schema_change: block`), Spice does not apply source schema changes automatically: the dataset stays healthy and continues serving queries using the schema registered at startup. To evolve the schema without restarting, set `on_schema_change` to one of:

- `append_new_columns`: add newly-appeared source columns; reject removals and incompatible changes.
- `sync_all_columns`: keep the registered schema synchronized with the source schema.
- `drop_and_recreate`: apply widening changes in place; otherwise drop and recreate the accelerated table (destructive, and only with `refresh_mode: full`).
- `fail`: fail the refresh when the source schema diverges.

Restarting the runtime also re-infers the schema from the source and re-initializes the accelerated table.

## What is Data-grounded AI?

Data-grounded AI anchors models in accurate, current, domain-specific data rather than relying solely on pre-trained knowledge. Spice unifies enterprise data across databases, data lakes, and APIs, dynamically incorporating real-world context at inference time. This helps minimize hallucinations, reduce operational risk, and build trust in AI by delivering reliable, relevant outputs.

## Where can I find examples and recipes?

The [Spice.ai Cookbook](https://github.com/spiceai/cookbook) provides quickstarts and examples demonstrating Spice capabilities, including federated queries, RAG, text-to-SQL, and more.