firebolt.io

# Firebolt

> The open-source analytical database for engineers — built for AI agents, real-time analytics, and efficient ELT, with the best price-performance in the market.

Firebolt is a high-performance, open-source analytical database (OSS available in preview today; GA launching August 2026). It runs as a single binary on your laptop and scales to hundreds of nodes and petabytes of data in your cloud — same binary, same behavior. It speaks the PostgreSQL SQL dialect, reads and writes Apache Iceberg and DuckLake, and serves queries over ADBC, JSON REST, and the Postgres wire protocol, with ACID transactions and snapshot isolation. Deploy it anywhere — single binary, Docker, Helm, or a Kubernetes operator — on bare metal, VMs, any cloud, or your own data center. Prefer not to self-host? Use the fully managed service or bring your own cloud (BYOC), available on AWS (GA), GCP (GA), and Azure (preview).

Install and run locally: `bash <(curl -s https://get.firebolt.io/)`

## Product

- [Firebolt: The Analytical Database for Engineers](https://www.firebolt.io/): Open-source analytical database built for AI agents, real-time analytics, and efficient ELT, with the best price-performance in the market.
- [Pricing](https://www.firebolt.io/pricing): Fully open source and free to self-host. Managed-service pricing is simple and transparent: per-second compute billing, scale to zero, and storage at object-store cost.
- [About](https://www.firebolt.io/about): Firebolt is a Postgres-compliant analytical database built for engineers — fast by default, simple by design, open by principle.
- [Security & Trust Center](https://www.firebolt.io/security): Firebolt's defense-in-depth architecture, compliance certifications (SOC 2, ISO 27001/27018, HIPAA), and security engineering practices.
- [FAQ](https://www.firebolt.io/faq): Frequently asked questions about Firebolt — performance, engines, pricing, security, integrations, and more.
- [Contact](https://www.firebolt.io/contact): Talk to Firebolt sales or support — book a call, email us, or reach out on social.
- [Careers](https://www.firebolt.io/careers): Join Firebolt and build the future of data and AI apps.

## Open Source

- [Firebolt Core on GitHub](https://github.com/firebolt-db/firebolt-core): Self-hosted, open-source Firebolt — free, forever. A single binary that runs from your laptop to a 256-node cluster with zero additional dependencies.
- [Introducing Firebolt Core — Self-Hosted Firebolt, For Free, Forever](https://www.firebolt.io/blog/introducing-firebolt-core): Firebolt's CTO and VP of Engineering on the launch of the self-managed, open-source version of Firebolt.
- [Firebolt MCP Server for LLM Integration](https://www.firebolt.io/blog/unlock-conversational-data-interaction-firebolt-mcp-server-for-advanced-llm-integration): Connect Firebolt to AI tools like Claude and Copilot via the Model Context Protocol to query data conversationally.

## Documentation

- [Firebolt Documentation](https://docs.firebolt.io/intro): Product docs — SQL reference, architecture, deployment modes, integrations, security, and operational guides.
- [Documentation (llms.txt)](https://docs.firebolt.io/llms.txt): LLM-friendly index of the full Firebolt documentation.
- [Changelog / Release Notes](https://docs.firebolt.io/reference/release-notes): Latest releases, including Iceberg catalogs, native JSON type, HNSW vector search indexes, and cross-database joins.

## Capabilities & Architecture

- [Querying Apache Iceberg with Sub-Second Performance](https://www.firebolt.io/blog/querying-apache-iceberg-with-sub-second-performance): How Firebolt's READ_ICEBERG provides low-latency access to Iceberg tables.
- [Efficient and ACID-Compliant Vector Search Indexes in Firebolt](https://www.firebolt.io/blog/technical-deep-dive-efficient-and-acid-compliant-vector-search-indexes-in-firebolt): HNSW vector search over float arrays for low-latency similarity search — relevant for AI/RAG workloads.
- [Live Engine Upgrades, Zero Downtime](https://www.firebolt.io/blog/live-engine-upgrades-zero-downtime-the-firebolt-method): Online scaling and upgrades using shadow clusters with no customer downtime.
- [Making a Query Engine Postgres Compliant](https://www.firebolt.io/blog/making-a-query-engine-postgres-compliant-part-i-functions): How Firebolt implements the PostgreSQL SQL dialect.
- [Making Firebolt Fast by Doing Practically Nothing](https://www.firebolt.io/blog/making-firebolt-fast-with-pruning): Pruning techniques that reduce scanned rows for sub-second performance.

## Customers

- [Customers](https://www.firebolt.io/customers): See how teams use Firebolt to power sub-second, customer-facing, and AI-driven analytics.
- [Credit Karma: 60 Billion Predictions Daily — Inside the Agentic Data Layer](https://www.firebolt.io/blog/60-billion-predictions-daily-inside-credit-karmas-agentic-data-layer): Building an agentic data layer for analytics at massive scale.
- [Similarweb: Sub-second analytics over 1+ trillion rows](https://www.firebolt.io/customers/how-similarweb-delivers-sub-second-analytics-over-1-trillion-rows-with-firebolt-over-aws): Raw data ready for dynamic querying at sub-second performance, no pre-processing required.
- [Sweet Security: Sub-second threat detection analytics](https://www.firebolt.io/customers/sweet): Ingests 1TB of security events daily across 350+ continuously updating tables under constant ingestion.
- [IQVIA: Life Science analytics](https://www.firebolt.io/customers/iqvia): Sub-second query performance for consistent BI access across hundreds of users.
- [Primer: Query performance with SQL only](https://www.firebolt.io/customers/primer): Eliminated 3–4 second delays in a customer-facing fintech product.
- [Bigabid: 400x query performance improvement](https://www.firebolt.io/customers/bigabid-slashes-latency-and-boosts-query-performance-400x-using-aws-and-firebolt): Reduced latency and boosted query performance 400x with Firebolt and AWS.

## Comparisons

- [Cloud Data Warehouse Comparison](https://www.firebolt.io/comparison): Side-by-side comparisons by architecture, scalability, performance, and cost.
- [Firebolt vs Snowflake](https://www.firebolt.io/comparison/firebolt-vs-snowflake)
- [Firebolt vs BigQuery](https://www.firebolt.io/comparison/firebolt-vs-bigquery)
- [Firebolt vs ClickHouse](https://www.firebolt.io/comparison/firebolt-vs-clickhouse)
- [Firebolt vs Databricks](https://www.firebolt.io/comparison/firebolt-vs-databricks)
- [Firebolt vs Redshift](https://www.firebolt.io/comparison/firebolt-vs-redshift)
- [Firebolt vs Athena](https://www.firebolt.io/comparison/firebolt-vs-athena)
- [Firebolt vs Druid](https://www.firebolt.io/comparison/firebolt-vs-druid)

## Optional

- [Blog](https://www.firebolt.io/blog): Engineering deep dives, customer stories, and The Data Engineering Show podcast. Full archive indexed at the blog and in the docs llms.txt.
- [Full comparison index](https://www.firebolt.io/comparison): All head-to-head data-warehouse and query-engine comparisons.
- [Privacy Policy](https://www.firebolt.io/legal/privacy)
- [Terms of Use](https://www.firebolt.io/legal/terms)
- [Master Subscription Agreement](https://www.firebolt.io/legal/master-subscription-agreement)
- [Cookies Policy](https://www.firebolt.io/legal/cookies)
- [Copyright Policy](https://www.firebolt.io/legal/copyright)