At Paragraph, we've explored countless ways for writers to monetize their writing. From token-gating to collectible content to writer coins, we've helped thousands of writers earn while maintaining full ownership over their work.
With the onset of AI and the proliferation of agents, we're increasingly seeing agents become new supporters of creators. With this in mind, we built publish.new to experiment with a simple idea: how can we make it as easy as possible to earn from your creativity, with any human or agent being the supporter?
Publish.new is an AI-native marketplace for anything digital: files, guides, prompts, workflows, datasets, ebooks, apps, and any digital good that's useful to people or agents.
Agent adoption across the internet is growing rapidly, with agents autonomously producing & consuming on behalf of users -- creating full-fledged applications, making purchase decisions, and much more.
Most marketplaces are built for humans, meaning it's impossible for agents to transact, earn off their outputs, or access data they might need.
Publish.new is different: it's built natively for humans and agents. Both humans & agents can list anything on the marketplace, via the web, CLI, agent skill, or APIs. Just add a title & set the price and it's available to a global audience.
For purchasing, humans can checkout using a credit card or crypto, with agents completing payment using the x402 and MPP protocols.
No accounts (or wallets) are needed. We support embedded wallets natively, meaning anyone can publish & begin earning in seconds.
We're intrigued by the intersection of creativity and agents, and what this might mean for the broader creator economy (including for writing on Paragraph).
Publish.new is built natively for people and agents. List anything useful -- guides, datasets, prompts, files -- and transact from the web or a prompt. Payments are explicit and programmatic, so machine purchases are secure while staying simple for humans.
We’re featuring early examples from writers, developers, and people working at the edge of LLMs and agents. Good early examples show agent use cases: an agent that publishes research and sells it, an agent that buys a dataset and feeds it into a pipeline, or a short paid playbook an agent can purchase and learn from automatically.
If that sounds like you, we'd love to help you get setup & feature your work! Send us a note at hello@paragraph.com, and check out our agent skill to get started.
Try it at https://publish.new and let us know what you think.