FTIR.fun FTIR.fun
Partner Integration Partners Only

FTIR.fun Partner API & MCP

This page is for partner companies who have been issued a partner API key. Two integration surfaces are available: a REST API for server-to-server automation, and a dedicated MCP server for AI agent frameworks.

🔗 Not a partner? General users and developers should use the standard integration page (/ai-integration/), which covers the public MCP, REST API, and Skills for all platforms.
1. Authentication 2. REST API 3. MCP Server 4. Rate limits 5. Quickstart 6. API Reference

1. Authentication

All partner endpoints require a Key created in the Partner API Key management page. Each institution must use a separate non-empty partner name; ordinary API Keys are rejected.

Billing is configured per partner Key: points mode deducts the configured number of points per analysis submission (zero is free), while subscription mode is unlimited until the configured expiry time. Status polling is not charged.

REST API — X-API-Key header

POST https://ftir.fun/v1/partner/compare-peaks
X-API-Key: ftir_your_key_here
Content-Type: application/json

MCP Server — Bearer token

{
  "mcpServers": {
    "ftirfun-partner": {
      "url": "https://ftir.fun/partner-mcp",
      "headers": {
        "Authorization": "Bearer ftir_your_key_here"
      }
    }
  }
}

2. REST API Endpoints

Base URL: https://ftir.fun/v1/partner/ — pass X-API-Key on every request.

POST /compare-peaks
Compare two FTIR peak lists. Returns similarity score, matched peaks, shift values, and coverage rates.
POST /explain-peaks-from-paper
Extract peak assignments from a paper by DOI or title. Returns paper-specific text excerpts and KG enrichment.
POST /verify-spectrum-against-paper
One-call: extract peaks from a paper, then compare them to your uploaded instrument file.
POST /verify-material
Test whether a spectrum matches a named material. Returns verdict (supported / conflicting), rank, and evidence.
POST /batch-analyze
Submit multiple spectra in one call. Peak-only batches of up to 10 may complete synchronously; larger or file batches are queued. Each sample is independent.
⟳ Queued jobs — poll GET /batch-analyze/{id}
POST /image-compare
Queue a comparison against a paper FTIR screenshot. A trained operator digitizes the paper figure; the returned estimate is not an SLA.
⟳ Async — poll GET /image-compare/{id}

3. MCP Server

Endpoint: https://ftir.fun/partner-mcp — 8 tools mirroring the REST API above.

compare_peak_listsCompare two FTIR peak sets explain_peaks_from_paperExtract assignments from paper by DOI verify_spectrum_against_paperOne-call paper vs. spectrum comparison verify_material_hypothesisTest spectrum against named material submit_batch_spectraSubmit batch analysis job get_batch_analysis_statusPoll batch job status submit_spectrum_image_comparisonSubmit image vs. spectrum comparison get_spectrum_image_comparison_statusPoll image comparison status

4. Rate Limits

EndpointPer minuteDaily cap
All REST endpoints60 requests—
Each MCP tool60 calls—

HTTP 429 on excess. The response includes a Retry-After header and retry guidance in recovery_suggestions. Limits are per API Key; the edge also protects the MCP path by client IP.

5. Python Quickstart

import requests, base64

API_KEY = "ftir_your_key_here"
BASE    = "https://ftir.fun/v1/partner"
HEADERS = {"X-API-Key": API_KEY, "Content-Type": "application/json"}

# 1. Compare two peak lists
r = requests.post(f"{BASE}/compare-peaks", headers=HEADERS, json={
    "peaks_a": [2915, 1720, 1460, 720],
    "peaks_b": [2918, 1718, 1465],
    "tolerance_cm1": 10,
})
print(r.json()["similarity_score"])   # 0.7144 for this example

# 2. Verify a material hypothesis from an instrument file
with open("sample.spa", "rb") as f:
    b64 = base64.b64encode(f.read()).decode()

r = requests.post(f"{BASE}/verify-material", headers=HEADERS, json={
    "material_name": "polypropylene",
    "file_base64": b64,
    "filename": "sample.spa",
})
print(r.json()["verdict"])   # "supported"

# 3. Extract peak assignments from a paper
r = requests.post(f"{BASE}/explain-peaks-from-paper", headers=HEADERS, json={
    "doi": "REPLACE_WITH_A_DOI_PRESENT_IN_THE_LOCAL_PAPER_CACHE",
    "compound": "PMMA",
})
for peak in r.json().get("ftir_peaks", []):
    print(peak["peak_cm1"], "—", peak.get("paper_assignment"))

6. Full API Reference

Interactive Swagger UI with live "Try it out" for every endpoint:

Open API Reference (Swagger UI) →

Machine-readable spec: /partner-openapi.json (import into Postman, Insomnia, or any OpenAPI-compatible tool)

Iesniegt pieprasījumu Veidlapa