Promptmetheus

Promptmetheus is a prompt engineering environment for writing, testing and refining the prompts behind LLM powered apps, agents, automations and workflows. It works like an integrated development environment for prompts rather than a chat window. Prompts are composed from structured text and data blocks, and each block can have variants so a writer can iterate quickly through versions. Test datasets hold dynamic context such as user input or retrieved data and are injected into prompts through data blocks. Custom automatic evaluators check completions against stated requirements, and completions can also be rated by hand, with results broken down by model and variant. Prompts can be tried against a catalog of more than 150 models from providers including Anthropic, Google DeepMind, OpenAI, Mistral, Perplexity, xAI, DeepSeek, Cohere, Groq and OpenRouter, and users can configure their own endpoints. Further features include project organization, project and prompt level variables, full version history with changelogs, inference cost estimates, real time sync across devices and team members, and export to text, CSV, Excel or JSON. It is aimed at prompt engineers, developers and teams who need prompts that hold up on output quality, cost and speed. The Single plan costs $29 per month for one user and includes a 7 day free trial. The Team plan costs $99 per month with three users included and $19 per month for each additional user. The product states it needs a screen of 12 inches or larger.