On October 9 we had 4,566 active remote listings. 1,181 of them have "engineer" or "developer" in the title. We searched those job descriptions for named AI tools and concepts. No model and no sampling: a plain text match over every description, with the query published alongside the data.
What employers name
More than half of the listings (636, or 54%) mention AI or machine learning somewhere. That number is less useful than it looks, because "we use AI to…" appears in plenty of company blurbs. Named tools tell you more:
| What the ad names | Listings | Share |
|---|---|---|
| "AI agent" or "agentic" | 253 | 21% |
| "LLM" | 187 | 16% |
| Claude or Anthropic | 101 | 8.6% |
| ChatGPT or OpenAI | 70 | 5.9% |
| Cursor | 64 | 5.4% |
| Claude Code, specifically | 57 | 4.8% |
| Copilot | 32 | 2.7% |
| MCP servers | 23 | 1.9% |
| LangChain | 22 | 1.9% |
| Vector databases | 14 | 1.2% |
| Gemini | 13 | 1.1% |
Two things stand out.
Agents beat any single tool. One listing in five says "AI agent" or "agentic". Employers care more about whether you can build or work alongside agents than about which chat window you use.
Coding tools are named as part of the job, not as perks. These are typical lines:
Use AI coding tools (e.g., Cursor, Claude Code, GitHub Copilot) as part of your daily workflow
Fluency with AI-assisted development tools (Cursor, Claude Code, MCP servers): this is part of how we work, not a…
Infrastructure that makes agents competent in our codebase (CLAUDE.md/AGENTS.md conventions…)
That last one is new: a team asking for someone to maintain the instruction files that their coding agents read.
Does naming AI move pay?
Only 266 of the 1,181 listings state a yearly USD range, so treat this as a signal rather than a market rate. We split those 266 by whether the description names an AI tool, LLM or agentic work:
| Group | Listings | 25th pct | Median | 75th pct |
|---|---|---|---|---|
| Names AI tools / LLMs / agents | 128 | $170,000 | $200,000 | $246,500 |
| Does not | 138 | $150,000 | $195,350 | $231,500 |
The medians are close, about $4,650 apart. The bottom of the range is where they differ: a quarter of the non-AI listings start at $150,000 or less, while the AI-naming quarter starts at $170,000. Listings that ask for AI fluency rarely sit at the low end.
What this means if you are job hunting
- Show one tool doing real work. "Familiar with ChatGPT" matches every CV. A repo where an agent writes and tests code against your own conventions file is something a reviewer can open.
- Learn the agent vocabulary. Agents, MCP and tool calling show up more often than any vendor name. If a listing says "agentic", expect questions about how you scope and check an agent's output.
- Read for the verb. "Use as part of your daily workflow" is a working requirement. "Exposure to" is a nice-to-have. Price your application accordingly.
How we counted, and what it misses
- Case-insensitive substring match on the full description of every active listing with "engineer" or "developer" in the title.
- "Cursor" includes Cursor's own job ads and mentions of it as a customer, so it over-counts a little. A spot check of "Claude" matches found every one referring to the AI tool.
- The pay comparison covers the 22% of listings that publish a yearly USD range, which skews toward US employers.
- This is our corpus of remote listings, not the whole job market.
We kept the exact queries and raw results, and we will share them with anyone who wants to recheck the numbers.
Applying to remote roles paid in USD? Get our free ATS-friendly remote CV template (.docx). Originally published on lokerdollar.com.
Top comments (0)