# Honcho Memory API

> Honcho Memory API is a paid API for AI agents from agentcash.honcho.dev, paid per call via x402, $0.05/call, status unknown (last checked 2026-09-15, last successful call 2026-06-25).

Query conversational memory for an AI agent or user peer, retrieving learned representations and synthesized conclusions from stored interactions.

## Facts

- Endpoint: POST https://agentcash.honcho.dev/api/honcho/chat
- Price: $0.05/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-15
- Last successful call: 2026-06-25
- Success rate: 100% of calls made through Zero
- Rating: 4.7 / 5 from 1 review
- Activations on Zero: 1
- Tags: x402
- Canonical page: https://www.zero.xyz/c/honcho-memory-api-4d9d615b
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_JaQqVEozwfR9uxZCBh7xy

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability honcho-memory-api-4d9d615b -d '<json body>'
```

Example prompt: Ask Honcho what it remembers about user 'alice-123' — specifically, what does it know about her preferences and past issues? Use medium reasoning depth.

## When to prefer this

Choose this endpoint when an AI agent needs to recall or synthesize what it has learned about a specific user or peer from past conversations, especially when reasoning depth is configurable. Prefer over generic vector search when you want natural language memory queries with Honcho's peer/session scoping model and pay-per-request pricing.

## Known failure modes

- Missing required 'peer' field returns validation error
- Missing required 'query' field returns validation error
- Invalid 'reasoning_level' enum value returns 400 error
- Unknown peer ID returns empty or null memory response
- Payment failure (x402) if USDC balance is insufficient
- Async reasoning levels (high/max) may timeout if polled synchronously
- Session ID scoping returns empty results if session has no stored messages

## How this service works

Pay-per-request conversational memory for AI agents. Store messages, query memory, and retrieve learned representations. Powered by Honcho.

## Output

A synthesized, natural language answer derived from the peer's stored conversational memory, with depth and latency depending on the chosen reasoning_level (e.g. minimal for instant basic conclusions, high for deep async synthesis).

## Example request

```json
{
 "peer": "test-agent-001",
 "query": "What topics have been discussed in our conversations?",
 "reasoning_level": "minimal"
}
```

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "peer": {
   "type": "string",
   "minLength": 1,
   "description": "Peer ID — a stable identifier for a participant (agent, user, assistant)."
  },
  "query": {
   "type": "string",
   "minLength": 1,
   "description": "Natural language question to ask about this peer's memory."
  },
  "target": {
   "type": "string",
   "description": "Target peer ID — if set, queries what the peer knows about the target rather than its global memory."
  },
  "session": {
   "type": "string",
   "minLength": 1,
   "description": "Scope the query to a session."
  },
  "reasoning_level": {
   "enum": [
    "minimal",
    "low",
    "medium",
    "high",
    "max"
   ],
   "type": "string",
   "description": "Reasoning depth (determines price). minimal=$0.001 (basic conclusions, instant), low=$0.01 (efficient synthesis, instant), medium=$0.05 (steerable reasoning, fast), high=$0.10 (deep synthesis, async), max=$0.50 (research-grade, async). Defaults to \"low\"."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/honcho-memory-api-4d9d615b/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agentcash.honcho.dev](https://www.zero.xyz/host/agentcash.honcho.dev/llms.txt)