- ·Up to $125/hr
- ·Evaluate AI responses on game-development scenarios
- ·1+ years full-time experience with Godot, LibreSprite, O3DE, or similar
All Gigs
- ·Full time W-2 role through an employer of record; 40 weekday hours weekly
- ·Maintainer or core contributor experience on a mature, widely used repository
- ·Build and test engineering tasks locally, including with Docker
- ·Up to $75/hr
- ·Evaluate AI on ML theory, optimization, and statistical inference
- ·PhD, doctoral student in ML, Stats, CS, or related
- ·Worldwide remote engagement for 40 weekday hours weekly
- ·2 or more years in ML systems or infrastructure
- ·Exclusive full time commitment with no other engagements
- ·Up to $105/hr, project-dependent
- ·4+ years in web development
- ·Remote, asynchronous, flexible hours
- ·4 or more years in cloud infrastructure or related engineering
- ·Production Kubernetes diagnosis and repair required
- ·Full time W2 role for 40 weekday hours weekly
- ·Up to $75/hr, remote and async with flexible hours
- ·Needs 4+ years in technical support or IT help desk
- ·Part-time, no minimum hours commitment
- ·Full time W2 employment in the California Bay Area
- ·40 hours weekly for an initial 6 month engagement
- ·Hybrid attendance; relocation costs are not covered
- ·Up to $105/hr, remote and fully asynchronous
- ·Needs 4+ years as a data scientist, analyst, or DBA
- ·Open to Bachelor's through Postdoc
- ·Talent Network OverviewThis is an open application for future contract opportunities that match
- ·Use Software Engineer ( - Native Platform & Integrations) expertise in structured AI review
- ·Produce written feedback on Software Engineer ( - Native Platform & Integrations) tasks
- ·Remote research projects offer rates up to $80 per hour.
- ·At least three years in security engineering or offensive security is required.
- ·Create security prompts and evaluate large language model responses.
- ·Full time W2 employment in the California Bay Area
- ·40 hours weekly for an initial 6 month engagement
- ·Hybrid attendance; relocation costs are not covered
- ·At least 10 years in industry, including 5 years of hands-on engineering
- ·At least 3 years leading engineering work lasting weeks or months
- ·Remote contract projects with no minimum hours or set schedule
- ·Remote US employment for 40 hours per week
- ·Create engineering tasks, reference solutions and evaluation criteria
- ·Requires 5+ years of industry experience and an engineering degree
- ·Project dependent pay up to $120 per hour
- ·Create and review software engineering challenges
- ·Remote contract with no minimum hours
- ·2+ years developing or optimizing NKI kernels on Trainium/Inferentia2
- ·Review CUDA→NKI fidelity and cross-hardware numerical correctness
- ·Assess SBUF/PSUM/HBM memory use, DMA, and NeuronCore performance
- ·Project dependent pay up to $60 per hour
- ·Create and review software engineering challenges
- ·Remote contract with no minimum hours
- ·3+ years in kernels across at least two of CUDA, Triton, NKI, or Pallas
- ·Assess numerical tolerances, runtime validity, and benchmark fairness
- ·Experience in at least three specified kernel task types required
- ·Review AI content about reservoir engineering, subsurface analysis and production
- ·Hands on experience with Petrel, ARIES, Techlog, Avocet or AVEVA PI
- ·Up to $85 per hour, typically 5 to 20 hours weekly with no minimum
- ·3+ years in application security, penetration testing, or vulnerability research
- ·Review Docker-based labs and functionality-plus-vulnerability verification
- ·CVE taxonomy and CVSS, CWE, and CAPEC knowledge required
Frequently asked questions
How much do AI training jobs pay?
Hourly rates listed on SignalGigs currently run from $3 to $400. General annotation and rating work sits at the low end; roles that need a professional background in law, medicine, finance or engineering sit at the high end. Each listing shows the pay range its source states, when it states one.
Do I need experience to get AI training work?
Entry-level annotation and rating work usually needs no prior experience beyond a qualification test. Higher-paying specialist roles ask for a degree or professional licence in the field. Each listing describes what it requires.
Which AI training platforms pay the most?
Pay depends more on the skill than on the platform. Specialist work in law, medicine, finance and advanced coding pays the most wherever it is posted; general annotation pays the least. Use the pay filter above to set a minimum hourly rate, or browse jobs grouped by pay at signalgigs.com/pay.
Can I do AI training work from anywhere?
Most listings are remote, but many are limited to particular countries for tax or employment-law reasons, and language roles need native-level fluency in that language. Check the location and requirements on each listing before applying.
How do I get started with AI training work?
Pick a listing that matches your skills and click Apply. You apply on the hiring company's own site, which usually starts with a qualification test or screening. SignalGigs itself needs no signup.
Is AI training a real job or a scam?
AI training is real paid work: companies building AI models pay people to write, rate, label and correct the data those models learn from. SignalGigs collects roles from the hiring platforms' own job boards, and each job page names the company. As with any contract work, be wary of anyone who asks you to pay a fee to apply.