Open data ยท CC BY 4.0September 2026

State of Agentic Hiring

As of September 2026, AgenticCareers.co tracks 1,734 open agentic roles across 449 companies. 43% disclose pay, at a median of $217,225. 744 roles opened in the last 30 days. The largest category is General AI (34.1%).

1,734
Open agentic roles
449
Companies hiring
$217,225
Median salary
744
Added in 30 days
01Compensation

Salary intelligence

Median $217,225 (p25 $181,750, p75 $254,950), USD annual, across 738 disclosed listings.

RoleMedian (USD)
AI Security$227,500
General AI$224,000
ML Engineer$198,650
Data Scientist$194,700
AI Agent Engineer$192,675

A median is published only where at least 30 listings disclose pay. Across our 2026 monthly snapshots, segments above that line moved under 5% month over month, while segments in the teens moved by as much as 29%, so a smaller sample would be noise reported as a finding. Held back this month, with their disclosed-pay counts: AI Solutions Engineer (22), RAG / Search (18), Forward Deployed Engineer (18), AI Infrastructure (17), AI Platform / SRE (16), AI Researcher (16), LLM Engineer (16), Generative AI (15), Conversational AI (15), AI Sales / GTM (7), AI Product Manager (6), Prompt Engineer (4), Computer Vision (3), NLP Engineer (3), AI Ops / DevOps (2), AI Safety (1), Robotics (1); and by seniority Junior (20), Executive (12), Intern (1).

02Momentum

Hiring velocity

744 new roles in the last 30 days (42.9% of the board). The categories adding roles fastest:

  • General AI+257
  • AI Agent Engineer+72
  • ML Engineer+62
  • AI Infrastructure+51
  • RAG / Search+41
03Landscape

Role & skill landscape

Where agentic hiring concentrates across role categories.

CategoryRolesShare
General AI59234.1%
AI Infrastructure18610.7%
AI Agent Engineer1448.3%
ML Engineer1287.4%
AI Solutions Engineer955.5%
AI Researcher854.9%
Data Scientist814.7%
RAG / Search583.3%
LLM Engineer533.1%
Generative AI523%
AI Security502.9%
Forward Deployed Engineer372.1%
AI Platform / SRE291.7%
AI Product Manager281.6%
Prompt Engineer261.5%
Computer Vision241.4%
Conversational AI171%
AI Sales / GTM140.8%
NLP Engineer100.6%
Robotics90.5%
AI Safety60.3%
AI Ops / DevOps60.3%
Full Stack AI40.2%
04Employers

Company leaderboard

449 companies hiring; the top 10 hold 45.1% of open roles.

  1. 01Databricks229
  2. 02Harvey98
  3. 03Anthropic96
  4. 04NVIDIA64
  5. 05Capital One62
  6. 06Scale AI56
  7. 07Adobe49
  8. 08Cloudflare46
  9. 09Brex43
  10. 10Ramp32
  11. 11Vercel28
  12. 12Twilio24
  13. 13Stripe23
  14. 14Ashby21
  15. 15MARLOWE COMPANIES INC.20
05Seniority

The market has almost no entry level

38 of 1734 live roles are junior or internship level.

LevelRolesShare
Mid67039%
Senior46727%
Staff36721%
Lead1428%
Executive503%
Junior292%
Intern91%

A normal engineering market has a base: more junior roles than staff roles, because companies expect to grow people. This one is inverted. Executive listings outnumber junior and internship listings combined, and mid level is not the middle of the distribution so much as the floor, the first level that arrives self-sufficient.

The likeliest explanation is capacity rather than hostility. Agent teams are mostly small, mostly new, and mostly assembled by moving experienced engineers across from other teams, which leaves nobody whose job is to grow a junior. That is a temporary condition, and the leading indicator to watch is internships: when that number moves, junior hiring follows about a year later.

Where the entry-level roles do exist is counterintuitive. They cluster with the smallest employers rather than the largest, which we cover in the entry-level analysis.

06Transparency

Who publishes pay, and who offers remote

44% of listings disclose a salary range, and the largest employers disclose least.

Employer segmentListingsDisclose payRemote
Ten largest employers77538%6%
Everyone else95948%27%
Employers with one open role34624%30%

The intuition is that large companies publish ranges because they have compensation bands and legal departments, while small teams stay vague. The data says the reverse, and because the ten largest employers hold a large share of all listings they pull the market-wide disclosure rate down on their own.

Remote runs the other way. The smallest employers advertise remote at roughly double the market rate, because a company hiring one person cannot restrict itself to a single metro area and still fill the role. The result is a coherent trade rather than a free lunch: the most accessible part of this market on location and seniority is also the least forthcoming about money.

LevelRolesRemote
Mid67022%
Senior46716%
Staff36713%
Lead14215%
Executive5020%

Remote share also falls as seniority rises. Both gradients trace to the same thing, which is that the highest-leverage work sits closest to the systems and to the people making decisions about them. Anyone optimising for remote and seniority at once is searching a genuinely small part of the market. The full breakdown by role category is in the remote gradient analysis.

07How it is built

Methodology

1,734 live roles from 15 sources. Salary: USD annual midpoint of disclosed range. A segment median is published only at n>=30 disclosed listings, the point at which month-over-month swing across our 2026 snapshots falls under 5%; smaller segments are named under salary.withheld with their sample size rather than estimated.. Every listing passed AgenticCareers' agentic relevance classifier; there is no non-agentic denominator.

08Reuse freely

Cite this dataset

Licensed CC BY 4.0. Cite as:

AgenticCareers.co, State of Agentic Hiring, September 2026. https://agenticcareers.co/state-of-agentic-hiring