Cross-provider arbitrage analytics for hosted inference
Cross-provider arbitrage analytics for hosted inference. Per-model: cheapest_paid, most_expensive_paid, spread_usd, savings_pct, median_paid_blended, fastest_tps, cheapest_with_tps, free_tier_offers. Plus provider_rollup (cheapest_count, top_tps_count, value_score 0-100) and top_arbitrage models sorted by savings_pct desc. Pairs with the free /api/inference-providers matrix. The "where to migrate workloads" call.
20000 (raw units)
price
1
calls / 30d
1
unique payers
2026-05-25
updated
Provider
tensorfeed.ai · discovered, not yet claimed by its owner
Payment (x402 accepts[])
[
{
"scheme": "exact",
"network": "eip155:8453",
"payTo": "0x549c82e6bfc54bdae9a2073744cbc2af5d1fc6d1",
"asset": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
"amount": "20000",
"maxTimeoutSeconds": 60
}
]Output schema
{
"bazaar": {
"info": {
"input": {
"method": "GET",
"queryFields": {
"family": {
"example": "Meta",
"type": "string"
},
"min_savings_pct": {
"example": 20,
"type": "integer"
}
},
"queryParams": {
"family": "Meta",
"min_savings_pct": 20
},
"type": "http"
},
"output": {
"example": {
"billing": {
"credits_charged": 1,
"credits_remaining": 49
},
"capturedAt": "2026-05-24T12:00:00Z",
"filter": {
"family": "Meta",
"min_savings_pct": 20
},
"matrix_last_updated": "2026-05-24",
"models": [
{
"cheapest_paid": {
"blendedPrice": 0.355,
"provider": "DeepInfra"
},
"cheapest_with_tps": {
"blendedPrice": 0.355,
"outputTPS": 170,
"provider": "DeepInfra"
},
"family": "Meta",
"fastest_tps": {
"outputTPS": 950,
"provider": "Groq"
},
"free_tier_offers": [],
"median_paid_blended": 0.385,
"modelId": "llama-4-scout",
"modelName": "Llama 4 Scout",
"most_expensive_paid": {
"blendedPrice": 0.425,
"provider": "Replicate"
},
"offer_count_paid": 6,
"paramsB": 109,
"savings_pct": 16.47,
"spread_usd": 0.07
}
],
"models_in_matrix": 3,
"ok": true,
"provider_rollup": [
{
"appearances_paid": 5,
"cheapest_count": 3,
"free_tier_count": 0,
"provider": "DeepInfra",
"top_tps_count": 0,
"value_score": 100
},
{
"appearances_paid": 3,
"cheapest_count": 0,
"free_tier_count": 0,
"provider": "Groq",
"top_tps_count": 2,
"value_score": 33
}
],
"top_arbitrage": [],
"tracked_providers": [
"Together AI",
"Fireworks",
"DeepInfra",
"Groq",
"OpenRouter",
"Replicate",
"Anyscale",
"DeepSeek",
"GitHub Models"
]
},
"schema": {
"properties": {
"billing": {
"properties": {
"credits_charged": {
"example": 1,
"type": "integer"
},
"credits_remaining": {
"example": 49,
"type": "integer"
}
},
"type": "object"
},
"capturedAt": {
"example": "2026-05-24T12:00:00Z",
"type": "string"
},
"filter": {
"properties": {
"family": {
"example": "Meta",
"type": "string"
},
"min_savings_pct": {
"example": 20,
"type": "integer"
}
},
"type": "object"
},
"matrix_last_updated": {
"example": "2026-05-24",
"type": "string"
},
"models": {
"items": {
"properties": {
"cheapest_paid": {
"properties": {
"blendedPrice": {
"example": 0.355,
"type": "number"
},
"provider": {
"example": "DeepInfra",
"type": "string"
}
},
"type": "object"
},
"cheapest_with_tps": {
"properties": {
"blendedPrice": {
"example": 0.355,
"type": "number"
},
"outputTPS": {
"example": 170,
"type": "integer"
},
"provider": {
"example": "DeepInfra",
"type": "string"
}
},
"type": "object"
},
"family": {
"example": "Meta",
"type": "string"
},
"fastest_tps": {
"properties": {
"outputTPS": {
"example": 950,
"type": "integer"
},
"provider": {
"example": "Groq",
"type": "string"
}
},
"type": "object"
},
"free_tier_offers": {
"type": "array"
},
"median_paid_blended": {
"example": 0.385,
"type": "number"
},
"modelId": {
"example": "llama-4-scout",
"type": "string"
},
"modelName": {
"example": "Llama 4 Scout",
"type": "string"
},
"most_expensive_paid": {
"properties": {
"blendedPrice": {
"example": 0.425,
"type": "number"
},
"provider": {
"example": "Replicate",
"type": "string"
}
},
"type": "object"
},
"offer_count_paid": {
"example": 6,
"type": "integer"
},
"paramsB": {
"example": 109,
"type": "integer"
},
"savings_pct": {
"example": 16.47,
"type": "number"
},
"spread_usd": {
"example": 0.07,
"type": "number"
}
},
"type": "object"
},
"type": "array"
},
"models_in_matrix": {
"example": 3,
"type": "integer"
},
"ok": {
"type": "boolean"
},
"provider_rollup": {
"items": {
"properties": {
"appearances_paid": {
"example": 5,
"type": "integer"
},
"cheapest_count": {
"example": 3,
"type": "integer"
},
"free_tier_count": {
"example": 0,
"type": "integer"
},
"provider": {
"example": "DeepInfra",
"type": "string"
},
"top_tps_count": {
"example": 0,
"type": "integer"
},
"value_score": {
"example": 100,
"type": "integer"
}
},
"type": "object"
},
"type": "array"
},
"top_arbitrage": {
"type": "array"
},
"tracked_providers": {
"items": {
"example": "Together AI",
"type": "string"
},
"type": "array"
}
},
"type": "object"
},
"type": "json"
}
},
"schema": {
"properties": {
"input": {
"additionalProperties": false,
"properties": {
"method": {
"enum": [
"GET"
],
"type": "string"
},
"queryParams": {
"properties": {
"family": {
"description": "Case-insensitive substring filter on model family (Meta, DeepSeek, Mistral, Alibaba, Microsoft, ...).",
"type": "string"
},
"min_savings_pct": {
"description": "Minimum savings_pct to include a model in the top_arbitrage array. Default 20.",
"maximum": 100,
"minimum": 0,
"type": "number"
}
},
"type": "object"
},
"type": {
"const": "http",
"type": "string"
}
},
"required": [
"type",
"method"
],
"type": "object"
},
"output": {
"properties": {
"example": {
"type": "object"
},
"type": {
"type": "string"
}
},
"required": [
"type"
],
"type": "object"
}
},
"required": [
"input"
],
"type": "object"
}
}
}Use it
curl
curl "https://tensorfeed.ai/api/premium/inference-providers/arbitrage" # -> 402 Payment Required, accepts[] lists how to pay # retry with a PAYMENT-SIGNATURE (or PAYMENT header) once paid
JavaScript
const res = await fetch("https://tensorfeed.ai/api/premium/inference-providers/arbitrage");
if (res.status === 402) {
const { accepts } = await res.json();
// pay one of accepts[] via an x402 client, then retry with the payment header
}Python
import httpx
res = httpx.get("https://tensorfeed.ai/api/premium/inference-providers/arbitrage")
if res.status_code == 402:
accepts = res.json()["accepts"]
# pay one of accepts[] via an x402 client, then retry with the payment headerMachine-readable
Everything on this page is also available as clean JSON at /resources/4191.json, and this resource appears in /discovery/resources and /discovery/search.