# ArbiPulse AI Inference Price Arbitrage

> ArbiPulse AI Inference Price Arbitrage is a paid API for AI agents from arbipulse.theaslangroupllc.com, paid per call via x402, $0.05/call, status unknown (last checked 2026-09-14).

Returns live pricing for every provider serving a given open-weights LLM model, including input/output cost per million tokens, blended 3:1 cost, cheapest-to-dearest spread multiple, and quantization caveats.

## Facts

- Endpoint: GET https://arbipulse.theaslangroupllc.com/api/inference-arb
- Price: $0.05/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/arbipulse-ai-inference-price-arbitrage-7c7c17ef
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_8ErCE6jz2OYQ5Z5CVhE0r

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 arbipulse-ai-inference-price-arbitrage-7c7c17ef
```

Example prompt: Show me all providers currently serving Llama 3 70B, with their input and output prices per million tokens, the blended 3:1 cost, the cheapest-to-dearest spread multiple, and any quantization caveats — I want to pick the cheapest option for my agent's inference spend.

## When to prefer this

Use this endpoint when an AI agent needs to minimize its own inference spend on a specific open-weights model, when comparing provider pricing before routing workloads, or when auditing inference cost arbitrage opportunities. Prefer over manual provider docs because it aggregates live pricing across 2-5+ providers simultaneously with a pre-computed spread multiple and blended cost metric.

## Known failure modes

- Model name not recognized or not tracked — returns empty provider list or 404-style error
- Model has only one known provider — spread multiple is 1x, caveat returned
- Pricing data temporarily stale — endpoint may return a freshness warning
- Payment not processed — x402 payment required error before data is returned

## How this service works

AI inference price arbitrage — every provider serving a given model with live input/output $/Mtoken, blended 3:1 cost, cheapest-vs-dearest spread multiple, quantization caveats. Same open-weights model spans 2-5x across providers; for agents that buy inference this is self-referential spend recovery. Live listings, no LLM.

## Output

A live, structured listing of every provider offering the requested open-weights model, each entry including: input $/Mtoken, output $/Mtoken, blended 3:1 cost, the spread multiple between the cheapest and most expensive provider, and any quantization or precision caveats. No LLM-generated content — data is pulled live.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "$schema": "https://json-schema.org/draft/2020-12/schema",
 "required": [
  "input"
 ],
 "properties": {
  "input": {
   "type": "object",
   "required": [
    "type",
    "method"
   ],
   "properties": {
    "type": {
     "type": "string",
     "const": "http"
    },
    "method": {
     "enum": [
      "GET",
      "HEAD",
      "DELETE"
     ],
     "type": "string"
    },
    "queryParams": {
     "type": "object",
     "properties": {
      "model": {
       "type": "string"
      }
     }
    }
   },
   "additionalProperties": false
  },
  "output": {
   "type": "object",
   "required": [
    "type"
   ],
   "properties": {
    "type": {
     "type": "string"
    },
    "errors": {
     "type": "object",
     "description": "Documented error responses, keyed by HTTP status code",
     "additionalProperties": {
      "type": "object",
      "required": [
       "description"
      ],
      "properties": {
       "example": {
        "type": "object"
       },
       "description": {
        "type": "string"
       }
      }
     }
    },
    "example": {
     "type": "object"
    }
   }
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "model": "meta-llama/llama-3.3-70b-instruct",
  "cheapest": "DeepInfra",
  "providers": [
   {
    "provider": "DeepInfra",
    "quantization": "fp8",
    "blended_usd_per_mtok": 0.2
   }
  ],
  "spread_multiple": 4.6
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/arbipulse-ai-inference-price-arbitrage-7c7c17ef/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from arbipulse.theaslangroupllc.com](https://www.zero.xyz/host/arbipulse.theaslangroupllc.com/llms.txt)