About the AI Use column: Over the last year we've seen a large increase in SDR software written with the help of AI coding tools such as Claude Code, OpenAI Codex, GitHub Copilot and Cursor. To help readers judge each program, we note what we know about its AI use. For open-source projects we checked the code repository for developer disclosures, AI co-author tags on commits, and AI agent instruction files such as CLAUDE.md or AGENTS.md. The 'AI Suspected' and 'vibe coded' ratings are our opinion only.

• AI Built The developer states, or the repository shows, that most of the code was written by AI.
• AI Assisted AI tools were used for part of the code or documentation, alongside substantial human development.
• AI Suspected Not disclosed, but in our opinion the project shows strong signs of being written mostly by AI.
• No AI Signs We checked the open-source code and found no indication of AI use.
• Pre-AI Last updated before AI coding assistants became widely available in mid-2022.
• Unknown Closed source with no disclosure, or not enough information to tell.
• Likely vibe coded In our opinion most of the code was generated by AI with limited human review or testing. This does not mean the software doesn't work, but extra caution and testing may be worthwhile.

★ Core Core programs are long-established staples. They have been in use for five years or more, are still maintained or still the standard tool for their job, and are widely used and recommended in the RTL-SDR community. They were chosen by a review panel using the popularity data below, and are listed first in each category.

Popularity: A rough guide to how widely used a program is, based on its GitHub stars (for open-source projects), how many of our posts mention it, how long it has been around, and how many people contribute to its code. Programs without a public code repository are rated on our coverage alone. Five bars marks the top 5% of this list. Hover over the bars to see the numbers.