# KARP by SoulDriver – Multi-Model Research Cards

> KARP by SoulDriver – Multi-Model Research Cards is a paid API for AI agents from agents.souldriver.com.au, paid per call via x402, $0.03/call, status unknown (last checked 2026-10-31).

Returns AI research cards synthesized by four models (GPT, Gemini, Grok, Claude) debating a topic, each card containing a conclusion, key insight, warnings, dated figures, consensus status, and open questions.

## Facts

- Endpoint: POST https://agents.souldriver.com.au/v1/ask?utm_source=zero.xyz
- Price: $0.03/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-31
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/karp-by-souldriver-multi-model-research-cards-88ab125f
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_osFbw4oR8OoqzQ6fUtbDg

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 karp-by-souldriver-multi-model-research-cards-88ab125f -d '<json body>'
```

Example prompt: Can you pull up KARP research cards on the current state of nuclear fusion energy — give me up to 6 cards covering conclusions, key insights, and any warnings about outdated data?

## When to prefer this

Choose this endpoint when you need a synthesized, multi-model research consensus on a topic rather than a single-model answer. It is especially valuable for policy, science, finance, or geopolitical topics where cross-model debate adds credibility. Prefer it over a simple web search when you want structured conclusions with explicit uncertainty flags, dated figures, staleness indicators, and open questions surfaced automatically. Use POST /v1/coverage first if unsure whether the topic is indexed.

## Known failure modes

- Query too vague or outside covered topics — low relevance scores or empty cards array
- max_cards out of range (must be 1–8) — validation error
- Topic not yet researched by KARP — coverage miss, recommend checking POST /v1/coverage first
- Stale cards returned for rapidly evolving topics — staleness flag will be true
- Network timeout given ~15s processing time — retry recommended
- Payment failure via x402 protocol — 402 response with payment details

## How this service works

Research cards from KARP, SoulDriver's multi-model research engine: GPT, Gemini, Grok and Claude debate a question and publish a conclusion. Each card: conclusion, key insight, warnings, dated figures, consensus, open questions, research date. Check your question first with the free POST /v1/coverage. Deepest on Alzheimer's, ageing biology, ancient scripts, AI governance, quantum computing. No wallet? Free research grant: POST /v1/grant. Usually 5-15 s.

## Output

A JSON object containing an array of research cards, each with: a ranked conclusion synthesized from multi-model debate, a key insight, warnings about limitations or outdated data, dated anchor figures with as-of dates, an age-in-days and staleness flag, open questions, a relevance score, a session ID, and the topic string. Also includes top-level relevance score and a list of related research topics with their session IDs.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "query": {
   "type": "string",
   "maxLength": 500,
   "description": "The question you are researching, in plain words. A specific question finds closer matches than a single keyword."
  },
  "intent": {
   "type": "string",
   "maxLength": 300,
   "description": "Optional: what you or your user are trying to find out, and what it is for (max 300 characters). Kept private; it helps decide what KARP researches next."
  },
  "max_cards": {
   "type": "integer",
   "default": 5,
   "maximum": 8,
   "minimum": 1,
   "description": "Research cards to return"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "cards": [
   {
    "date": "2026-04-06",
    "rank": 1,
    "stale": true,
    "topic": "AI governance 2026: latest regulatory developments, EU AI Act enforcement updates, and Australia's AI safety framework progress",
    "anchors": [
     {
      "as_of": "2026-04-07",
      "label": "ISO 42001 Certification Cost",
      "value": "$85,000 - $650,000"
     }
    ],
    "age_days": 193,
    "warnings": [
     "Agentic AI may operate outside existing legal frameworks."
    ],
    "relevance": 0.994,
    "conclusion": "A bifurcated global landscape: the EU has paused high-risk enforcement pending harmonised standards, while Australia enforces AI obligations through existing privacy law.",
    "session_id": "karp_1775506768",
    "key_insight": "The 40% grey zone in high-risk AI categorisation reflects a failure of categorical logic, not just technical standards.",
    "open_questions": [
     "How will legal frameworks adapt to agentic AI systems?"
    ]
   }
  ],
  "query": "AI governance in 2026",
  "top_relevance": 0.994,
  "related_research": [
   {
    "date": "2026-04-10",
    "topic": "Timeline of the US federal-state AI regulatory collision",
    "session_id": "karp_1775819207"
   }
  ]
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/karp-by-souldriver-multi-model-research-cards-88ab125f/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agents.souldriver.com.au](https://www.zero.xyz/host/agents.souldriver.com.au/llms.txt)