# FoundryNet Inference

> FoundryNet Inference is a paid API for AI agents from foundrynet-inference-production.up.railway.app, paid per call via x402, $0.02/call, status unknown (last checked 2026-09-14).

LLM inference proxy with data-enriched analysis and predictive intelligence, backed by 17 data sources, with MINT-attested outputs gated via x402 micropayments

## Facts

- Endpoint: POST https://foundrynet-inference-production.up.railway.app/v1/chat
- Price: $0.02/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/foundrynet-inference-c490bcfc
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_noA2D7t2pInpHnVXWKGea

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 foundrynet-inference-c490bcfc -d '<json body>'
```

Example prompt: Ask FoundryNet Inference to analyze the current macroeconomic outlook for emerging markets using its 17 data sources and give me a data-enriched, MINT-attested summary — use the default model with a max of 512 tokens.

## When to prefer this

Choose FoundryNet Inference when you need LLM completions that go beyond a single model's training data by pulling from 17 live data sources, and when output attestation (MINT) is required for auditability or trust. Prefer this over bare LLM APIs when you need enriched, verifiable, predictive intelligence rather than raw text generation. Useful in agentic pipelines where downstream systems need to verify that an AI response was legitimately produced and data-grounded.

## Known failure modes

- 402 Payment Required if no valid x402 payment or fnet_ Forge key is provided
- 400 Bad Request if the messages array is missing or malformed
- Invalid model name results in model-not-found error
- Upstream LLM provider unavailability causes 503 or timeout
- Rate limiting or quota exceeded on data enrichment sources returns partial or degraded response
- max_tokens set too low may truncate the response mid-generation

## How this service works

LLM inference proxy + data-enriched analysis + predictive intelligence. 17 data sources. MINT-attested outputs. x402-gated (Solana/Base USDC); an fnet_ Forge key bypasses.

## Output

A chat completion response generated by the configured LLM, enriched with insights from up to 17 external data sources, accompanied by a MINT attestation that cryptographically verifies the output's provenance and integrity.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "messages"
 ],
 "properties": {
  "model": {
   "type": "string",
   "description": "LLM model to use"
  },
  "system": {
   "type": "string",
   "description": "Optional system prompt"
  },
  "messages": {
   "type": "array",
   "description": "Chat messages array"
  },
  "max_tokens": {
   "type": "integer",
   "description": "Optional max output tokens"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/foundrynet-inference-c490bcfc/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from foundrynet-inference-production.up.railway.app](https://www.zero.xyz/host/foundrynet-inference-production.up.railway.app/llms.txt)