OnlyPrompts

OnlyPrompts is an AI chat bot tool. OnlyPrompts combines a 150,000 prompt library for 37,000+ workplace tasks with side-by-side prompt testing and an in-app assistant that runs GPT-4, Claude 3 Sonnet and Claude 3 Haiku. It suits professionals who want curated, reusable prompts and a record of which versions produce the best output.

OnlyPrompts is a prompt library and testing workspace aimed at people who use large language models for everyday professional work but do not want to craft every instruction from scratch. The catalog holds more than 150,000 prompts covering over 37,000 tasks, grouped by function such as marketing, sales, finance, administration, customer service, data analysis, research and development, human resources and information technology. Each entry is written as a reusable template, so the research and development example asks the model to act as an aerospace expert and simplify technical content for a named audience at a stated level of understanding, with placeholders the user fills in. In practice the product does two things beyond browsing. First, a configuration toolbox lets users run prompts side by side and compare the outputs, adjust the wording, and save the versions that performed best. Second, an AI Assistant runs prompts directly inside the platform, with the ability to switch between GPT-4, Claude 3 Sonnet and Claude 3 Haiku without leaving the page. Credits act as the internal currency for those test runs and model chats, which keeps experimentation and generation in one place instead of spread across separate subscriptions and browser tabs. The typical user is a marketer, analyst, HR generalist, IT staffer or founder who already pays for one or more chat models and wants more predictable results from them. The prompt templates are described as compatible with multiple large language models, so a team can keep its preferred model and still reuse the same instruction set. Output varies by model, and the platform does not claim to tune prompts automatically per model, so comparison testing remains a manual exercise. Against general prompt marketplaces, the differentiator is the combination of a large categorized library with native A/B testing and in-app model access. Against just writing prompts by hand in ChatGPT, the trade-off is a subscription or one-time purchase in exchange for a curated starting point and a saved record of what works. The catalog leans toward business and productivity tasks, so users looking mainly for creative image or video prompting will find thinner coverage than the occupational categories suggest.