Find remote Prompt Engineer jobs.
Shape what models say.

An AI prompter job is a role where you write and test the instructions that make an AI model give useful answers. Job titles vary: prompt engineer, AI prompter, AI writer or AI trainer. Pure prompt-writing roles are rare; most listings combine prompting with evaluation, RAG pipelines or product work, and AI training platforms hire prompters by the hour to write and rate model answers. The roles on this page and the AI trainer page cover both.

Create a free profile and browse the prompt engineer roles on this page. Describe your LLM systems in detail: the use case, the model stack, the evaluation approach, and the production results. Companies hiring serious prompt engineers search profiles and look for engineers who describe systems, not just skills. A profile that shows evaluation rigor and production experience says more than a list of AI tools.

It depends on seniority, location and the company, so we do not quote a typical figure. Many listings on this page show the pay range the employer published: compare those directly, and check whether the range is base salary or includes equity.

Evaluation methodology (RAGAS, human eval, custom metrics), RAG pipeline design (chunking, embedding models, vector databases), prompt versioning and A/B testing, model selection judgment (when to use which model and why), cost and token optimization, and guardrail implementation. Also valuable: LangChain, LlamaIndex, LangSmith for observability, and OpenAI Assistants API with function calling. Describe your evaluation approach first: it is the most differentiating signal in this role.

Yes, for normal use. Creating a profile, browsing jobs and applying are free, up to 20 applications a day. Credits are optional: they pay for applications beyond that limit or to put an application at the top of the list. No placement fee when you get hired, no share of your salary, no exclusive contract.

Yes, but the skill set is evolving. The engineers building the most durable careers in prompt engineering are those who treat it as a full engineering discipline: evaluation design, retrieval architecture, cost optimization, and reliability engineering. Pure prompt writing without this broader skill set is more vulnerable to automation. The combination of strong evaluation thinking, RAG depth, and LLM system design is what these listings ask for.

The most effective portfolios include 2–3 documented LLM systems: the use case (what the system does and who uses it), the model and orchestration stack, the evaluation framework you built (how you define and measure output quality), a specific challenge you solved (retrieval quality, hallucination, cost, latency), and the outcome. GitHub repositories with prompt libraries, evaluation scripts, or RAG pipeline code are strong additions. Public demos or open-source LLM tools you have contributed to are also compelling.

The transition requires building two things: an evaluation framework and a production RAG pipeline. Build a system where you can systematically measure prompt quality across many test cases, not just "this prompt looks better than that one." Then build a RAG pipeline end-to-end: document ingestion, chunking, embedding, retrieval, generation, and evaluation. Document both projects thoroughly with the design decisions and results. These two components demonstrate that you treat prompting as engineering, which is what companies are actually hiring for.

Yes. RemoteAI lists freelance and contract prompt engineering roles in addition to full-time positions. Freelance prompt engineering typically involves building evaluation frameworks, RAG pipelines, or prompt optimization systems for specific projects. Filter by job type to see freelance and contract options. Building a clear portfolio of documented LLM systems is particularly important for freelance work, as clients need to assess your capability from your profile alone.