MotherDuck—the serverless cloud data warehouse built on DuckDB—has quietly made a significant pricing change. The $25/month Lite plan that small teams relied on is gone, replaced by a free tier with severe limitations. The Business plan jumped from $100 to $250/month. No blog post. No announcement. Just a pricing page update sometime between December 2025 and February 2026.
If you’re a small data team, startup, or solo practitioner who was using MotherDuck’s affordable plans, this change affects you directly. Here’s what happened, who it impacts, and what your options are.
What Exactly Changed
MotherDuck’s pricing has shifted three times since its general availability launch in June 2024. We tracked every change using Web Archive snapshots of their pricing page.
The Full Timeline
June 2024 — GA Launch:
- Standard Plan: $25/month
- Included 100 GB storage + 100 CU (Compute Unit) hours per month
- Pay-per-query billing ($0.0000694/CU second) for usage beyond included amounts
- Unlimited users
- 30-day free trial
February 2025 — First Restructuring:
- Standard Plan replaced by Lite ($25/month) and Business ($100/month)
- Compute and storage separated from the platform fee — no longer included in base price
- Storage: $0.08/GB/month
- New per-instance compute options (Pulse, Standard, Jumbo) alongside pay-per-query
- Existing Standard Plan customers auto-transitioned to Business by May 1, 2025
January/February 2026 — Current Pricing:
- Lite Plan: Dropped to $0/month (free) — but severely gutted
- Business Plan: Increased to $250/month — a 2.5x jump from $100
- Storage reduced to $0.04/GB/month
- New compute tiers: Mega ($12/hr) and Giga ($36/hr) added
Side-by-Side Comparison
| Feature | GA Standard (Jun 2024) | Lite (Feb 2025) | Business (Feb 2025) | Lite (Now) | Business (Now) |
|---|---|---|---|---|---|
| Platform Fee | $25/mo | $25/mo | $100/mo | $0 | $250/mo |
| Storage Included | 100 GB | None | None | 10 GB | None |
| Compute Included | 100 CU hrs | None | None | 10 CU hrs (Pulse) | None |
| Max Users | Unlimited | 5 | Unlimited | 3 | 10 |
| Compute Types | Pay-per-query | Pulse + Standard | Pulse + Standard + Jumbo | Pulse only | All (Pulse—Giga) |
| Read Scaling | No | No | Yes | No | Yes |
| SLA | None | Standard | Priority | Community | 99.9% |
| Backup Retention | N/A | N/A | N/A | 1 day | 90 days |
The numbers tell the story. In 18 months, the minimum viable paid plan went from $25/month with 100 GB storage and 100 CU hours included to $250/month with nothing included.
The “Missing Middle” Problem
The real issue isn’t that the Business plan costs $250. Enterprise features like 99.9% SLA, 90-day backups, and priority support justify a premium. The problem is the gap between free and paid.
The free Lite plan gives you:
- 10 GB of storage
- 10 hours of Pulse compute per month
- Maximum 3 users
- Pulse instances only (the lightest compute tier)
- 1-day backup retention
- Community support only
For many small teams, 10 GB of storage and 10 hours of Pulse compute is barely enough for a proof of concept. The moment you outgrow these limits—which happens quickly with any real analytical workload—your only option is the $250/month Business plan.
There is no $50 plan. No $100 plan. No “Team” tier for growing startups. It’s free or $250+.
Who This Affects
Solo practitioners and hobbyists can still use the free tier for experimentation. If your data fits in 10 GB compressed and your queries stay under 10 Pulse hours/month, you won’t pay anything.
Small data teams (3-10 people) are hit hardest. The free tier caps at 3 users and offers only Pulse compute. A team of 5 analysts running moderate queries will immediately need the $250/month Business plan—plus compute and storage costs on top.
Startups have a partial escape route. MotherDuck’s Startup Program offers a 50% discount on annual Business plan contracts. That brings it down to ~$125/month, but requires an annual commitment with a minimum $6,000/year contract.
Enterprise teams are the target audience for this pricing change. The $250/month Business plan with 99.9% SLA, unlimited service accounts, all compute tiers, and 90-day point-in-time restore is competitive for organizations already paying thousands for Snowflake or BigQuery.
What MotherDuck Is Signaling
This pricing change reveals a clear strategic shift. MotherDuck is pivoting from a developer-friendly, low-cost alternative to Snowflake toward an enterprise-focused cloud data warehouse. The free tier exists to maintain the developer on-ramp, but the monetization is aimed at teams with budget.
This follows an industry pattern. ClickHouse Cloud raised prices ~30% in January 2025. Cloud data platform companies across the board are moving from growth-mode pricing to sustainability-mode pricing as venture capital tightens and investors demand revenue.
The hybrid execution model—where local DuckDB queries cost nothing and only cloud queries are billed—remains MotherDuck’s strongest differentiator. But if your workload requires collaboration, persistent cloud storage, or more than 3 users, you’re paying the full Business tier.
Alternatives for Affected Users
If the new pricing doesn’t work for your team, here are practical options.
1. Self-Hosted DuckDB + Object Storage
DuckDB itself is free, MIT-licensed, and incredibly fast. You can run it locally or on a VM with data stored in S3, GCS, or Azure Blob Storage as Parquet files.
Pros: Zero cost beyond infrastructure. Full control. No vendor lock-in.
Cons: Single-user (no concurrent writes). No built-in collaboration. You manage backups, security, and availability yourself. Definite documented their migration from Snowflake to self-hosted DuckDB, achieving 70%+ cost reduction—but they had to build workarounds for DuckDB’s single-writer limitation.
2. BigQuery (Google Cloud)
BigQuery’s on-demand pricing charges $5/TB scanned with 1 TB of free queries per month. There’s no platform fee.
Pros: Truly pay-per-query. 1 TB free/month covers many small workloads. Full SQL. Scales to petabytes.
Cons: Google Cloud lock-in. Less efficient for small-to-medium data where DuckDB excels. Query costs can spike unpredictably on large scans.
3. ClickHouse Cloud
ClickHouse Cloud offers a Basic tier with pay-as-you-go pricing. Instances auto-idle when not in use.
Pros: Sub-second query performance. Strong compression (10-40x). Open-source foundation. Good for real-time analytics and log analytics use cases.
Cons: Learning curve if you’re coming from DuckDB/SQL. Column-oriented storage requires different schema design thinking. Our managed ClickHouse service can help teams who need ClickHouse but don’t want to manage it.
4. PostgreSQL + Columnar Extensions
If your team already runs PostgreSQL, extensions like pg_analytics (from ParadeDB) or Citus add columnar storage and analytical query capabilities.
Pros: No new database to learn. Leverages existing PostgreSQL infrastructure. ACID transactions. Well-understood operational model.
Cons: Slower than purpose-built OLAP engines for large analytical queries. Not a replacement for petabyte-scale analytics.
5. Continue Using MotherDuck’s Free Tier Strategically
Use MotherDuck’s free tier for cloud collaboration and sharing, but do heavy computation locally with DuckDB. MotherDuck’s hybrid execution model means queries on local data don’t incur cloud compute charges. Store reference datasets in MotherDuck (within 10 GB), do the heavy analytics locally, and only use cloud compute for shared dashboards or API access.
Cost Comparison for a Typical Small Team
To put this in perspective, here’s what a small analytics team (5 analysts, 50 GB data, 100 compute hours/month) would pay across platforms:
| Platform | Monthly Cost (Estimate) |
|---|---|
| MotherDuck (Jun 2024) | ~$25 (most usage within included limits) |
| MotherDuck (Now) | $250 + ~$60 compute + |
| BigQuery (on-demand) | $0-$25 (if queries stay under 1 TB scanned/month) |
| ClickHouse Cloud (Basic) | ~$50-$150 (depending on compute hours) |
| Self-hosted DuckDB + S3 | ~$5-$20 (VM + storage costs only) |
| Snowflake | ~$300-$500 (compute credits + storage) |
MotherDuck went from being the most affordable managed option to pricing itself alongside Snowflake for small teams. The value proposition now depends entirely on whether you need MotherDuck’s specific features (hybrid execution, DuckDB compatibility, collaboration) enough to justify the premium.
What Should You Do?
If you’re on the free tier and it works: Stay. 10 GB and 10 Pulse hours is fine for POCs, personal projects, and light exploration.
If you’re a startup: Apply for the Startup Program for the 50% discount before committing.
If you’re a small team that outgrew the free tier: Evaluate BigQuery or ClickHouse Cloud first. Both offer more granular pay-as-you-go pricing without a $250/month platform fee.
If you’re enterprise: The Business plan at $250/month is actually competitive when you factor in the 99.9% SLA, priority support, and all compute tiers. Compare it against your current Snowflake or BigQuery bill.
If you love DuckDB but not the price: Run DuckDB locally with data in S3/GCS Parquet files. You lose collaboration features but keep the query performance at near-zero cost.
Need Help Choosing the Right Analytics Stack?
Navigating pricing changes, migrations, and platform decisions is exactly what we do. Our team helps data teams evaluate, migrate, and optimize their analytics infrastructure—whether that’s managed ClickHouse, data analytics pipelines, or cloud-native data warehouse orchestration with Airflow.
- Vendor-neutral guidance on the right analytics platform for your workload
- Migration support from MotherDuck, Snowflake, or BigQuery to cost-effective alternatives
- Managed operations for ClickHouse, PostgreSQL, and data pipeline infrastructure
Talk to our data engineering team about your analytics stack →