GitHub Copilot: what developers actually say
We read every Hacker News comment from Nov 2025 – Nov 2026 that gave a real opinion on GitHub Copilot. 501 of the 659 opinions come from people who say they use it themselves.
Across 659 opinions from 535 authors, GitHub Copilot holds a 42% approval rating. The last 90 days sit lower at 38%. Code quality is the heaviest complaint area at 93 negative mentions to 38 positive, with reliability at 54 to 14 and cost at 80 to 60. Autonomy is the only aspect that leans positive, 24 to 13, and 40 developers named Claude Code as the tool they preferred instead, against 18 who preferred Copilot over Claude Code. Monthly approval since July 2026 has ranged from 17% to 61% on 25 to 29 opinions per month, a small sample.
Pick it if
- Want autocomplete inside VS Code or Visual Studio
- Already pay for Copilot and want chat and CLI in the same plan
- Work in an IDE and prefer chat over a terminal UI
Look elsewhere if
- Care about code quality on larger changes
- Need reliable multi-step agent runs
- Want predictable model costs or generous usage limits
Strengths and weaknesses
Each row counts opinions that mentioned the topic. Bars to the right of the line are praise, bars to the left are complaints.
How opinion has moved
What developers say
Praise
Complaints
Summaries are paraphrased. Each links to the original comment.
GitHub Copilot compared
Best GitHub Copilot alternatives, according to developers who switched →
FAQ
Is GitHub Copilot worth it?
Approval is 42% across 659 opinions, and 38% in the last 90 days. Cost draws 80 negative mentions against 60 positive, and usage limits add 31 negative to 20 positive. It fits developers already paying for it who mainly want autocomplete in an IDE.
Is GitHub Copilot better than Claude Code?
The data leans toward Claude Code. 40 developers said they preferred Claude Code instead, against 18 who preferred Copilot over Claude Code. Copilot's clearest edge is autonomy, where positive mentions outnumber negative ones 24 to 13.
How does Copilot compare on cost?
Cost has 80 negative mentions against 60 positive. Developers report that Anthropic models through Copilot can run up a large bill, and one spent about $500 in credits on a single effort. Usage limits add 31 negative mentions to 20 positive.