Machine Learning salaries in San Francisco.
What Machine Learning roles in San Francisco pay, from live listings on JobsRadar.
Compare with anotherMedian salary
$232,000
per year · annualized USD
Typical range
$195k – $287k
25th–75th percentile
Disclosing pay
50.2%
1,627 of 3,242 roles
Live roles
3,242
open right now
Salary trend
median · annualized USDSalary distribution
annualized USD · disclosed rolesHighest-paying jobs
Research Engineer, Machine Learning (RL Velocity)
Research Engineer/Research Scientist, Pre-training
Performance Engineer
Research Engineer, RL Engineering
Staff Research Engineer, Discovery Team
Research Engineer, Code RL (Reinforcement Learning)
Research Engineer, Safeguards Labs
Research Engineer, Knowledge Team
Top companies hiring
| Company | Median | Roles |
|---|---|---|
|
Anthropic
|
$362,500 | 135 |
OpenAI
|
$337,500 | 128 |
Okta
|
$230,500 | 97 |
Databricks
|
$218,450 | 94 |
Baseten
|
$247,500 | 67 |
| SA Scale AI | $243,000 | 62 |
Pinterest
|
$256,377 | 52 |
Socure
|
$173,750 | 30 |
See every Machine Learning role
Browse the live listings, or save this search and get alerted the moment a new one posts.
How much do Machine Learning roles in San Francisco pay?
The median salary for Machine Learning roles in San Francisco is $232,000 per year. Most pay falls between $195,000 (25th percentile) and $287,000 (75th percentile), with the broader 10th–90th percentile band running from $161,500 to $348,102.
These figures are computed live from 3,242 active Machine Learning roles in San Francisco on JobsRadar, of which 1,627 (50.2%) publish a salary range. Pay is normalized to annualized USD so roles can be compared on equal terms; the highest-paying roles listed above show their original posted currency. Numbers refresh every few hours as new roles are posted and older ones close.
Employers currently hiring well-paid Machine Learning roles in San Francisco include Anthropic.
Salary transparency varies by employer and region, so the disclosing share above reflects only roles that publish pay — not every open position. Use it as a directional benchmark rather than an exact offer.