Polar Signals Cloud
polarsignals.com
- Free trial
- Linux
- Web
- API
- Self-hosted
What Polar Signals Cloud is
Polar Signals Cloud is a hosted service for storing and analyzing continuous profiling data. Its agent collects profiles from a node without code instrumentation, then sends them to the cloud. Documentation lists support for C and C++, Rust, Go, Ruby, Python, PHP, Java, Clojure, Kotlin, JVM workloads, Graal Native, Node.js, and .NET; additional product details list Erlang and Julia. The Pro tier provides CPU, memory, and GPU profiling for any language. Profiles can be explored with flame graphs, function-level filters, and Prometheus-style selectors with custom dimensions. An MCP integration lets AI tools query profiling data for bottlenecks and optimization opportunities. The Profiler extension displays inline metrics in VS Code and compatible editors such as Cursor, Windsurf, and Antigravity. Agents run in Kubernetes, Docker, ECS, or bare-metal environments. A 14-day trial includes Pro features with rate limits and 14-day retention. Pro costs $50 a month, billed per 100 million samples, with 30-day retention.
Who it suits
Polar Signals Cloud suits engineering teams that need continuous profiling across cloud, container, or bare-metal deployments. It also fits developers who want profiling metrics in editors or queried through AI tools.
Good to know, and what to watch
- Zero-instrumentation profile collection
- CPU, memory, and GPU profiling
- Flame graphs and function-level filtering
- MCP access for AI profiling queries
- Supports Kubernetes, Docker, ECS, and bare metal
- No free plan
- Trial retention is 14 days
- Pro uses usage-based billing
- BYOC requires contacting sales
Plans
Every plan and priceEnterprise
Free trial
Pro
14-day trial
The verdict
Polar Signals Cloud combines hosted profile storage with broad language and deployment support. Its editor extension and MCP integration add ways to inspect profiling data, while retention and usage billing vary by plan.
What the maker publishes
- AI integration
- Polar Signals Cloud provides an MCP integration that lets AI tools query profiling data and help identify bottlenecks and optimization opportunities. source
- Collection
- Its zero-instrumentation approach collects profiling data by deploying an agent on a node and sending the data to Polar Signals Cloud. source
- Company
- Polar Signals describes itself as a remote-first software company founded in 2020. source
- Deployment environments
- The agent supports Kubernetes, Docker, ECS, and bare-metal environments, according to the product site. source
- GPU profiling
- The product site describes per-kernel timing, CPU stack traces, and CUDA Graph support for GPU profiling. source
- IDE integration
- The Polar Signals Profiler extension displays profiling metrics as inline source-code annotations in VS Code and compatible editors including Cursor, Windsurf, and Antigravity. source
- Languages
- The documentation lists support for C/C++, Rust, Go, Ruby, Python, PHP, Java/Clojure/Kotlin/JVM, Graal Native, Node.js, and .NET. source
- Product
- Polar Signals Cloud is a hosted service for storing and analyzing continuous profiling data. source
- Profiling types
- The Pro tier includes CPU, memory, and GPU profiling for any language. source
- Retention limits
- The trial includes 14-day retention and Pro includes 30-day data retention. source
- Security and compliance
- The Pro tier lists SOC 2 Type 2 compliance and SSO support; the SSO documentation says credentials are stored encrypted in a Vault instance and login is served over HTTPS. source
- Self-hosted deployment
- Enterprise customers can request BYOC, and documentation describes deployment on Kubernetes with S3-compatible storage and PostgreSQL; BYOC is not self-serve and requires contacting sales. source
- Support
- The Pro tier includes Discord and Slack support, while Enterprise adds premium support, dedicated Slack Connect, and the ability to page on-call. source
- Visualization and queries
- The product site describes flame graphs, function-level filtering, and Prometheus-style selectors with custom dimensions. source
Checked against the maker's pages on 7 Oct 2026.