ClickHouse started 2026 with major announcements: a $400 million Series D round valuing the company at $15 billion, the acquisition of Langfuse for LLM observability, and the launch of a native Postgres service. Here’s everything you need to know about ClickHouse in 2026.
$400M Series D at $15B Valuation
On January 16, 2026, ClickHouse announced a $400 million Series D funding round, more than doubling its valuation to $15 billion in less than a year.
Investment details:
| Metric | Value |
|---|---|
| Funding amount | $400 million |
| Valuation | $15 billion |
| Lead investor | Dragoneer Investment Group |
| ARR growth | 250%+ year-over-year |
| Cloud customers | 3,000+ |
Participating investors include Bessemer Venture Partners, GIC, Index Ventures, Khosla Ventures, Lightspeed Venture Partners, T. Rowe Price Associates, and WCM Investment Management.
Recent customer wins include Capital One, Lovable, Decagon, Polymarket, and Airwallex, joining existing customers like Meta and Tesla.
Langfuse Acquisition: Entering LLM Observability
ClickHouse acquired Langfuse, the open-source LLM observability platform, to expand into AI monitoring and evaluation.
Langfuse by the numbers:
- 20,000+ GitHub stars
- 26M+ monthly SDK installations
- 6M+ Docker pulls
- Trusted by 19 Fortune 50 and 63 Fortune 500 companies
Marc Klingen, Langfuse CEO, explained the acquisition: “We built Langfuse on ClickHouse because LLM observability and evaluation is fundamentally a data problem.”
This positions ClickHouse to compete in the growing AI observability market alongside tools like Datadog and New Relic.
Native Postgres Service Launch
ClickHouse announced an enterprise-grade Postgres offering integrated with its analytics platform. This enables:
- Unified workloads - Transactional and analytical data in one platform
- 100x faster analytics - Via automatic data sync from Postgres to ClickHouse
- Native CDC - Change Data Capture built-in
- Managed infrastructure - Partnership with Ubicloud
This directly challenges the need for separate OLTP and OLAP databases, offering a unified solution for teams running Postgres for transactions and needing fast analytics.
ClickHouse 25.12 Release Highlights
The ClickHouse 25.12 release delivers significant performance improvements:
75x Faster Lazy Reading
A reimagined lazy reading execution model delivers 75x faster query performance for specific workloads, particularly benefiting dashboard and interactive analytics use cases.
Faster Top-N Queries
New data skipping indexes optimize top-N queries, reducing scan times for time-series and log analytics workloads.
Improved Join Performance
The DPsize join reordering algorithm has been enhanced, automatically optimizing complex multi-table joins.
Data Lake Integrations
Expanded support for:
- Apache Iceberg
- Delta Lake
- Popular data catalogs
chDB v4.0: Zero-Copy Pandas Integration
chDB v4.0 achieves true zero-copy integration with Pandas DataFrames:
import chdb
import pandas as pd
# Direct memory sharing - no serialization
df = pd.DataFrame({'id': range(1000000), 'value': range(1000000)})
result = chdb.query("SELECT * FROM df WHERE value > 500000", output_format="DataFrame")Performance improvement: Queries that previously took 30 seconds now complete in under 1 second by eliminating serialization and implementing direct memory sharing between ClickHouse and NumPy.
Other Acquisitions in 2026
ClickHouse has made three acquisitions:
| Company | Focus Area |
|---|---|
| Langfuse | LLM observability |
| HyperDX | Open-source observability platform |
| PeerDB | Postgres CDC and replication |
These acquisitions strengthen ClickHouse’s position in observability and database integration.
2026 Roadmap
According to the January 2026 newsletter, ClickHouse plans to deliver:
ClickStack improvements:
- Fully managed ClickStack experience
- AI-powered notebooks
- Anomaly detection
- Out-of-the-box integrations
- Opinionated defaults for observability
Performance and enterprise features:
- Faster dashboards, searches, and histograms for high-volume environments
- RBAC (Role-Based Access Control) for administrators
- Audit logging
Deeper cloud integration across all offerings.
Community Highlights
Notable Contributions
lgbo from BIGO contributed multiple performance optimizations:
- Reduced memory usage for window functions
- Improved hash table iteration
- Optimized CROSS JOINs
- New
stringComparefunction for lexicographic comparisons
Interesting Projects
| Project | Description |
|---|---|
| DoomHouse | 3D game engine rendering graphics entirely in ClickHouse SQL |
| clickgraph | Graph query engine using Cypher syntax |
| clickspectre | Analyzer tracking actual table usage patterns |
Migration Success Stories
WKRP: TimescaleDB to ClickHouse
WKRP migrated their RuneScape tracking plugin from TimescaleDB to ClickHouse with impressive results:
| Data Type | Before | After | Reduction |
|---|---|---|---|
| Location data | 4.28 GiB | 592 MiB | 87% |
| XP tracking | 872 MiB | 168 MiB | 81% |
They also replaced Apache Flink with ClickHouse’s native Kafka integration for real-time streaming.
Why Companies Build on ClickHouse
Luke Reilly’s analysis explains why seven major companies (Uber, Cloudflare, Instacart, GitLab, Lyft, Microsoft, Contentsquare) build identical four-layer abstraction stacks over ClickHouse:
- Ingestion layer - Kafka, Flink, or native ClickHouse Kafka engine
- Storage layer - ClickHouse MergeTree tables
- Query layer - ClickHouse SQL with materialized views
- API layer - Custom services or ClickHouse HTTP interface
This architecture enables sublinear scaling where engineering headcount grows with data volume rather than user count.
What This Means for Your Stack
ClickHouse’s 2026 moves signal several trends:
For observability teams: The Langfuse and HyperDX acquisitions position ClickHouse as a unified backend for logs, metrics, traces, and now LLM observability. Consider ClickHouse-based stacks like SigNoz for cost-effective observability.
For data teams: The native Postgres service eliminates the need for separate OLTP and OLAP databases. If you’re running Postgres + a data warehouse, evaluate the unified offering.
For AI teams: Langfuse integration means native LLM observability without additional tooling. This could simplify AI monitoring stacks significantly.
Related Resources
- ClickHouse vs Snowflake 2026
- ClickHouse vs BigQuery 2026
- ClickHouse vs Elasticsearch 2026
- All-in-One Observability Stack 2026
- How Tasrie IT Services Uses ClickHouse
Need Help with ClickHouse?
Evaluating ClickHouse for analytics or observability? Our team helps organizations design and implement ClickHouse-based data platforms.