AI Visibility Report: AI Coding Tools
30 brands, 10 AI models, 4,200+ mentions across 1,000 responses. Engineers reach for Cursor, Claude Code, and Codex. AI search recommends Tabnine. Two different markets, one buyer pool.
Key Findings
GitHub Copilot leads AI visibility in the coding-assistant space with 55.3% across 10 AI models, with Cursor close behind at 55.0%. Tabnine — a tool that peaked in 2022 — sits at #4 thanks to training-data echo. Claude Code and Codex, the agents engineers actually reach for in 2026, rank #3 and #8 because half the AI models tested don't know they exist.
AI Visibility Rankings — Top 15 of 30 Brands
- GitHub Copilot — 55.3% (10/10 models)
- Cursor — 55% (10/10 models)
- Claude Code — 33.7% (8/10 models)
- Tabnine — 29.7% (10/10 models)
- Windsurf — 28.8% (10/10 models)
- Amazon Q Developer — 19.5% (10/10 models)
- Sourcegraph Cody — 17.9% (10/10 models)
- Codex — 16.4% (9/10 models)
- Replit — 16.3% (10/10 models)
- Augment Code — 14.4% (10/10 models)
- Aider — 13.5% (9/10 models)
- Continue — 13% (10/10 models)
- Qodo — 11.9% (9/10 models)
- JetBrains AI Assistant — 8.7% (10/10 models)
- Bolt.new — 7.9% (8/10 models)
Methodology
100 prompts were run across 10 AI models (ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Grok, Mistral, Qwen, Llama, Command R) to measure how 30 ai coding tools brands appear in AI responses. The weighted visibility score accounts for recommendation frequency, ranking position, and model coverage.
Report Sections
- The Thesis
- What We Measured
- How We Measure AI Visibility
- AI Visibility Rankings
- What the Models Reveal About Themselves
- Where the Models Disagree
- Three Asymmetries Worth Acting On
- Head-to-Head: Who Wins When AI Decides
- The Visibility Landscape, Mapped
- The Sources AI Trusts
- What Kind of Content AI Actually Cites
- Do Models Push Their Parent's Product?
- How Different Model Categories See the Market
- Where Models Disagree Most
- The Playbook
- Methodology
Download the full PDF report
