The same intelligence.
Inside your agent.
Choose your connection. Ask your question.
Add Agentery to your assistant.
https://agentery.com/api/mcp
claude mcp add --transport http agentery https://agentery.com/api/mcp
{
"mcpServers": {
"agentery": {
"url": "https://agentery.com/api/mcp"
}
}
}https://agentery.com/api/mcp
{
"servers": {
"agentery": {
"type": "http",
"url": "https://agentery.com/api/mcp"
}
}
}codex mcp add agentery --url https://agentery.com/api/mcp
gemini mcp add --transport http -s user agentery https://agentery.com/api/mcp
curl -s https://agentery.com/api/mcp \
-H 'content-type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'Add the server
Use the command or settings for your client below. Public reads need no account or key.
Check the tools
Your client lists the tools automatically. Start with search_providers or get_group_fair_price.
Ask a pricing question
Your assistant can search, inspect profiles, compare products and check fair prices.
A tested request.
get_group_fair_price for a customer-support chatbot. Checked on 10 October 2026: HTTP 200 in 585 ms.
curl -s https://agentery.com/api/mcp \
-H 'content-type: application/json' \
-d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"get_group_fair_price","arguments":{"query":"customer support chatbot"}}}'{
"jsonrpc": "2.0",
"id": 2,
"result": {
"content": [
{
"type": "text",
"text": "<JSON below, as a string>"
}
]
}
}
// result.content[0].text, parsed (excerpt):
{
"status": "ok",
"group_id": 54,
"group": "Customer support automation",
"listings": 44,
"priced_listings": 42,
"fair_price_by_tier": {
"individual": {
"median_monthly_usd": 21.35,
"typical_range_usd": [
10,
40
],
"listings": 17
},
"pro": {
"median_monthly_usd": 59,
"typical_range_usd": [
29,
139
],
"listings": 25
},
"team / small business": {
"median_monthly_usd": 199.5,
"typical_range_usd": [
90.71,
321.76
],
"listings": 20
},
"enterprise": {
"median_monthly_usd": 999.5,
"typical_range_usd": [
535.29,
1092.2
],
"listings": 8
}
},
"typical_payment_model": "subscription",
"method": "Fair price is given per buyer tier: the median of the group's listings' entry monthly price within that tier (individual, pro, team / small business, enterprise), with the typical range (middle half), from current published plans converted to USD. A tier's median is shown only when at least 3 listings price that tier.",
"as_of": "2026-10-10",
"currency": "USD (converted from published currencies)",
"note": "Medians of current published prices in the group. Missing prices are unknown, not zero; contact-sales plans have no numeric price."
}Start with a useful question.
“What do comparable customer-support chatbots charge per month for a team?”
“Find AI meeting-notes tools for a sales team, then compare the three best on price and billing terms.”
“A vendor quoted $400 a month for an AI SDR agent for a 20-person team. Is that in line with comparable services?”
“How have agent prices moved over the last 90 days, and which buyer tier changed most?”
“What are the alternatives to Chatbase, and which of them publish a lower comparable price?”
“Does Chat Thing have a free plan, and is its top plan bought at checkout or through sales?”
Tool reference
Choose a service
search_providers → get_provider_profile → compare_providers
Check a price
get_group_fair_price · price_benchmark · market_report
Follow the market
get_price_index · get_price_index_history
Research a question
request_research · get_research
Parameters, responses and errors for every tool are in the documentation. The live tools/list is authoritative.
What is a fair market price?
The median monthly price of comparable services: products in the same group, for one buyer tier, in USD. Each answer shows the buyer tier, currency, billing basis, sample size and what was excluded. How it is calculated.
What will my agent receive?
Structured product identities, observed pricing, billing terms and source evidence. A missing price is returned as unknown, never as zero. The website, MCP, REST and A2A use the same handlers, so the answers match.
If a call fails, the error says how to fix it: a code, a fix with the corrected call or next step, and whether a retry can help.
What we record
When your assistant calls a tool, we record the tool, the arguments it sent, a summary of our answer, a one-way hash of the IP address and its country. We never see your wider conversation, chat history, memory or files.
Full request captures are kept for 90 days and only the operator can see them. Query text may be sent to the AI model providers we use, for processing only. Privacy policy
What we promise
- Every public tool and page is free and needs no account.
- A missing price is shown as unknown, never as zero, and a contact-sales price is not treated as a number.
- Methodology changes and data corrections are recorded in the changelog.
- Agentery does not host, run or resell the products it lists.
Partner-only tools (the workflow ranker) need a key; nothing else does.
Limits
MCP: 30 tool calls per minute per session or IP (60 after five correlated outcome reports in 30 days). REST and A2A: 60 requests per minute per IP. No account or key is needed for public reads.