Roles at frontier labs · Research specialties
Evaluations Researcher and Engineer
22 open · October 2, 2026
Evaluation teams decide what a model can do and whether a change made it better: benchmarks, graders, model judges, and the statistics that say whether a gain is real.
- Evals and benchmarks 100%
- How language models work 64%
- Agents and tool use 64%
- Post-training and fine-tuning 45%
- Python 45%
- Production systems 41%
taught here course coming not taught
The numbers
- Open roles now at 10 labs
- 22
- Median base pay middle half $310k to $440k, 11 U.S. bands
- $353k
- State no degree 5% require a PhD outright
- 82%
- Years asked, median of the 5 that state a minimum
- 3
Who is hiring
10 labs are hiring
In the labs' own words
In the labs' own words
“Have hands-on experience with LLMs, RL, RLHF/RLAIF, post-training, evals, graders, synthetic data, model training, coding agents, tool-using agents, or production ML systems.”
“You have improved LLM-powered, agent-powered, or AI-product systems through evals, feedback loops, data curation, prompting, model adaptation, or model selection.”
“Experience in implementing or developing benchmarks for large language models and multimodal models (e.g., vision-language, audio, video, browser agents)”
Your way in
Three doors to this role
- Door 1
Finishing a degree
18% mention publications
The way in is a program: 44 open now
- Door 2
A software engineer
82% state no degree
What to add
- Door 3
A PhD
18% ask for distributed training
Engineering at scale
Asked often
The short answers
How much does a evaluations researcher and engineer at a frontier lab make?
The median U.S. base salary across 11 open evaluations researcher and engineer postings that publish a band is $353k; the middle half runs from $310k to $440k. Equity and bonus are not included.
Do you need a PhD to be a evaluations researcher and engineer?
82% of the 22 open postings state no degree at all, 14% mention a PhD or master's, and 5% require a PhD outright.
What skills do evaluations researcher and engineer postings ask for?
Most often: evals and benchmarks (100%), how language models work (64%), agents and tool use (64%), post-training and fine-tuning (45%), python (45%).
Which labs are hiring evaluations researcher and engineers?
Right now: OpenAI (4), Cohere (3), Meta (Superintelligence Labs and FAIR) (3), Anthropic (2), Cursor (Anysphere) (2), Reflection AI (2), and 4 more.
Prepare for it here
A free account opens the courses and saves your progress.
Create your free accountNewsletter
One lesson by email each week.
One lesson a week. Free, and it stays free.
The board
Open roles, newest first
Each title opens the lab's own posting. By family: software engineering at labs · research scientists and engineers
- Engineering Manager, Evals ↗
Cursor (Anysphere) · Software engineering at labs · No degree stated
- Software Engineer, Agent Evaluation and Quality ↗
Cursor (Anysphere) · Software engineering at labs · No degree stated
- Member of Technical Staff, Evaluation Execution ↗
METR · Research scientists and engineers · No degree stated
- Cyber Evaluations Engineer ↗
Anthropic · Software engineering at labs · No degree stated · $300k to $405k base
- Research Engineer, Post-Training Model Evaluations ↗
Anthropic · Research scientists and engineers · No degree stated · $500k to $850k base
- Researcher, Evals ↗
Cartesia · Research scientists and engineers · No degree stated · $220k to $350k base
- Member of Technical Staff, Data Analysis and Evaluation ↗
Cohere · Research scientists and engineers · No degree stated
- Member of Technical Staff, North Modeling (Evals) ↗
Cohere · Research scientists and engineers · No degree stated
- Senior Research Scientist, Model Evaluation ↗
Cohere · Research scientists and engineers · No degree stated
- Research Engineer, Evals Infra - MSL FAIR Foundations ↗
Meta (Superintelligence Labs and FAIR) · Research scientists and engineers · No degree stated
- Research Engineer, Privacy Evals - Meta Superintelligence Labs ↗
Meta (Superintelligence Labs and FAIR) · Research scientists and engineers · Master's, PhD or equivalent
- Research Engineer, Safety Evaluation ↗
Meta (Superintelligence Labs and FAIR) · Research scientists and engineers · PhD required
- Agent Post-Training, Frontier Evals and Environments Research ↗
OpenAI · Research scientists and engineers · No degree stated · $380k to $500k base
- Backend Software Engineer (Evals) ↗
OpenAI · Software engineering at labs · No degree stated · $266k to $445k base
- Research Engineer, Frontier Evals & Environments ↗
OpenAI · Research scientists and engineers · No degree stated · $295k to $380k base
- Researcher, Agent Safety, Training and Evaluations ↗
OpenAI · Research scientists and engineers · No degree stated · $380k to $500k base
- Member of Technical Staff - Evaluations ↗
Reflection AI · Research scientists and engineers · No degree stated
- Member of Technical Staff - Research Software Engineer - Safety Evaluations Infrastructure ↗
Reflection AI · Research scientists and engineers · No degree stated
- Research, Post-Training Evals ↗
Thinking Machines Lab · Research scientists and engineers · PhD or equivalent experience · $350k to $475k base
- Software Engineer, Evaluation Platform / Infra ↗
Thinking Machines Lab · Software engineering at labs · Bachelor's or equivalent · $300k to $350k base
- Member of Technical Staff - Evaluation Infrastructure ↗
xAI · Research scientists and engineers · No degree stated · $180k to $440k base
- Software Engineer - Evals ↗
xAI · Software engineering at labs · No degree stated · $175k to $275k base
