TIOBE Index February 2026: Python Is Slipping, C++ Overtakes Java, and R Won't Stay Quiet
The February 2026 Rankings
The TIOBE Index for February 2026 is out, and the top 10 looks different than it did a year ago:
- Python — 21.81% (down)
- C — 11.05% (up)
- C++ — 8.55% (up)
- Java — 8.12% (down)
- C# — 6.83% (stable)
- JavaScript — 2.92% (stable)
- Visual Basic — 2.85% (stable)
- R — 2.19% (up from 15th)
- SQL — 1.93% (down)
- Delphi/Object Pascal — 1.88% (down)
Just outside the top 10: Perl at 11th (1.67%), up from 30th a year ago.
Python Is Still #1 — But the Gap Is Closing
Python leads by more than 10 points over second-place C. That's still dominant. But the trend line is worth watching.
In July 2025, Python peaked at 26.98%. Six months later, it's at 21.81% — a drop of over 5 percentage points. That's not noise. That's a trend.
TIOBE CEO Paul Jansen attributed the shift to "several more specialized or domain-specific languages gradually gaining ground at Python's expense."
What's happening isn't that developers are abandoning Python. It's that the ecosystem is fragmenting. As AI, data science, systems programming, and web development each mature, developers are reaching for tools that are purpose-built for their domain rather than defaulting to Python for everything.
Python became the universal language by being good enough at everything. But "good enough" is losing ground to "built for this."
C++ Overtakes Java
For the first time in the modern TIOBE era, C++ has overtaken Java — sitting at 8.55% versus Java's 8.12%.
This isn't a sudden shift. It's the culmination of years of C++ modernization (C++20, C++23) that made the language significantly more approachable while retaining its performance advantages. Meanwhile, Java has been steady but hasn't generated the same excitement.
Key drivers for C++:
- Game engines (Unreal, custom engines) remain C++ dominant
- Embedded systems and IoT continue to grow
- AI inference at the edge often requires C/C++ for performance
- C++20/23 features (modules, coroutines, ranges) reduced the "C++ is painful" narrative
Java isn't dying — enterprise backends, Android (though Kotlin is eating into that), and massive legacy codebases ensure its relevance. But the momentum has shifted.
R's Quiet Comeback
The most surprising mover: R climbed from 15th to 8th in just 12 months.
R was supposed to be the language Python killed. For years, data scientists migrated from R to Python, and the narrative was that R was a niche academic tool that couldn't compete with Python's broader ecosystem.
So what changed?
- Specialization wins — R's statistical computing capabilities remain unmatched for specific tasks. As data science matures, practitioners are choosing the best tool for the job rather than the most popular one
- Tidyverse ecosystem — R's data manipulation and visualization libraries (ggplot2, dplyr, tidyr) are genuinely best-in-class for exploratory data analysis
- AI can write R — Ironically, AI coding assistants have made R more accessible. You don't need to memorize R syntax when Copilot or Claude can write it for you
- Bioinformatics and academia — These fields never left R, and they're growing rapidly
Perl's Return From the Dead
Perhaps the most unexpected entry: Perl at 11th, up from 30th a year ago. It even briefly hit 10th place in September 2025.
Perl was widely considered a dead language — a relic of the CGI web era. But it turns out there's a massive amount of Perl infrastructure still running in production across telecom, finance, and sysadmin tooling. As those systems need maintenance and the pool of Perl developers shrinks, demand (and search volume, which TIOBE measures) is rising.
It's less a comeback and more a reminder that "dead" languages run more critical infrastructure than most developers realize.
What About Rust and Go?
Neither Rust nor Go appear in the TIOBE top 10, despite being among the most-loved languages in developer surveys. This highlights a key limitation of the TIOBE Index: it measures search engine query volume, not developer satisfaction or adoption velocity.
Rust and Go are growing rapidly in specific domains — Rust in systems programming and security-critical applications, Go in cloud infrastructure and DevOps. But they're still niche compared to the broad reach of Python, C, or Java.
The JetBrains State of Developer Ecosystem and Stack Overflow Survey tell a different story, where Rust and Go consistently rank highest on "want to learn" and "most admired" lists. TIOBE captures breadth; surveys capture depth.
What This Means for Developers
Don't Chase Rankings
The TIOBE Index is a temperature check, not a career guide. Python dropping from 27% to 22% doesn't mean you should stop learning Python. It means the industry is diversifying.
Specialization Is Winning
The era of "one language for everything" is fading. The best developers in 2026 are polyglot — they pick the right language for the domain rather than forcing one language to do everything.
Legacy Languages Still Matter
Perl at 11th, COBOL still hanging on, Delphi in the top 10 — these are reminders that production systems outlast hype cycles. There's real career value in knowing languages that others have written off.
Watch the Systems Languages
C and C++ are both trending up. As AI moves from cloud to edge, from training to inference, performance-critical languages are becoming more relevant, not less.
The Bottom Line
Python is still king, but the court is getting crowded. The February 2026 TIOBE Index tells a story of specialization over generalization — languages built for specific domains are reclaiming territory from the universal default.
For developers, the takeaway is simple: learn the language that solves your problem best, not the one with the highest ranking.