DeepNLP AI Agent Marketplace

Overview

The directory spans more than 30 categories, including sales, human resources, finance, legal, writing, and coding. Its entries cover both public repositories such as GitHub and Hugging Face and cloud AI agent services. That breadth helps people survey different kinds of agents in one place, though a directory is a starting point for choosing software rather than a substitute for evaluating an individual offering.

Search is available on the web and through a Search API and MCP servers. Users can submit metadata through the website, curl, or Python, with approval required before entries appear in categories. The ai_agent_marketplace Python package and API-key submission flow give developers a practical route to contribute, while the approval step means publishing is not immediate.

OneKey offers access to commercial and non-commercial MCPs on credit-based pricing, with discounts of up to 50%. Credits can also be earned by writing reviews, uploading content, and sharing. These options broaden the surrounding ecosystem, but they are distinct from the core directory’s search function.

Key features

• Large, varied index: More than 10,000 listings across 30-plus categories make it suited to broad discovery, including coding agents, AI employees, deep research agents, and image generators.
• Programmatic discovery and submission: Search API and MCP access help developers work with the index, while website, curl, and Python submission routes support publishing metadata. Approval is required before listings appear.
• Deployment and monetization: Every plan includes agent hosting and monetization; paid tiers raise hosting capacity and add production-grade rate limits and quota. The free tier is therefore useful for a small start, not equivalent to the paid production limits.
• OneKey MCP access: Credit-based access covers commercial and non-commercial MCPs, with discounts of up to 50%; review, content, and sharing activity can earn credits.

Pricing

DeepNLP uses a freemium model. Free costs 0.00 USD per month and includes exploration across 30-plus categories, one agent deployment hosting, a live subdomain, and agent monetization. It is a sensible way to explore the directory or host a single agent, but the one-hosting cap and lack of production-grade rate limits and quota make it a limited choice for heavier deployment needs.

Pro Monthly costs 9.90 USD per month and includes five deployment hostings, live subdomains, monetization, and production-grade rate limits and quota, alongside directory access. It fits users who need several hosted agents and prefer monthly billing. Pro Annually costs 99.00 USD per year and includes eight hostings and the same production-grade limits and quota. It suits users willing to pay annually for more hosting capacity; compared with the monthly plan, it offers three additional hostings, though the annual commitment is less flexible.

Platforms

DeepNLP is available on the web and through an API. The marketplace also supports access through MCP servers, and the publishing channels include the ChatGPT App Store, Google App Store, and MCP. This mix is relevant to both people browsing in a browser and developers integrating search or submitting metadata programmatically.

Who it's for

Choose DeepNLP to compare agents across categories, search an index through API or MCP, or publish agent metadata. It is also a reasonable entry point for a creator who needs one hosted agent and a live subdomain at no charge. Look elsewhere if your priority is a particular agent product rather than cross-category discovery, or if you need production-grade rate limits and quota without paying for Pro.

Pros and cons

• Pro: More than 10,000 entries across 30-plus categories give cross-category researchers a broad discovery pool.
• Pro: Web, Search API, MCP, and Python submission options accommodate both casual browsing and technical workflows.
• Pro: The free plan includes one hosting, a live subdomain, and monetization, so trying deployment does not require an upfront subscription.
• Con: Free hosting is capped at one agent and lacks the production-grade rate limits and quota offered on Pro, limiting its fit for production workloads.
• Con: Submissions need approval before appearing in categories, so contributors cannot count on immediate publication.

Alternatives

For a broader directory comparison, see AI Agent Marketplaces. Choose Poe instead if you want an AI collaboration product available across Android, iOS, macOS, web, and Windows, with a free plan that includes a 200k-token context length. GPT Store is a better fit for signed-in users seeking access to custom GPTs, with limited messages, image creation, and voice chats on its free plan.

AgenMart suits users focused on AI text generation and content optimization, with a free plan that includes 10,000 words of monthly usage and API access. Pick AGENTIS if its API and web platforms and free Explorer plan better match your agent needs. Relevance AI may suit a small team seeking a web-based AI workforce, with its free plan capped at 100 credits per day, one user, and 10MB of knowledge.

Microsoft 365 Copilot Agent Store is the alternative to consider if you need an agent store across Microsoft 365 platforms and are prepared for paid plans. AI Agent Store is another paid web-based option. AiCrate is worth considering for a web-based, credit-priced option, with paid plans beginning at 9.99 USD per month for 500 credits.

Verdict

DeepNLP is a strong choice for people who want a wide-ranging agent directory with API and MCP search and a route to submit metadata. Its main advantage is breadth paired with multiple access paths; its main drawback is that hosting is capped on Free and production-grade limits require Pro. Choose it for discovery and modest hosting, but look elsewhere if you need a different focused product or production capacity without a paid plan.