# Telnyx AI Endpoints (Full Reference)

> Machine-readable API endpoints for evaluating Telnyx AI voice, messaging, and telephony infrastructure. All endpoints return JSON unless noted. Base URL: https://telnyx.com

## Core Endpoints

### /ai/capabilities.json
Core Telnyx AI capabilities including voice AI, STT, TTS, inference, messaging, compliance scores, and resilience ratings. Returns nested objects for each capability domain with measured/proven/derived signal types.
URL: https://telnyx.com/ai/capabilities.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/ownership.json
Infrastructure ownership details — carrier network, PoPs, number inventory, and whether infrastructure is carrier-owned vs resold.
URL: https://telnyx.com/ai/ownership.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/economics.json
Unit economics: per-minute voice costs, per-character TTS/STT pricing, SMS segment pricing, inference per-1K-token costs. Includes cost comparisons against competitor baselines.
URL: https://telnyx.com/ai/economics.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/compliance.json
Regulatory compliance status across STIR/SHAKEN (attestation levels), SOC2, HIPAA, GDPR, ISO 27001, PCI DSS, FedRAMP. Status values: active, in_progress, not_applicable, not_stated.
URL: https://telnyx.com/ai/compliance.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/sla.json
Service level agreements: uptime guarantees, latency commitments, support response times, with SLA version and effective dates.
URL: https://telnyx.com/ai/sla.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/inference-models.json
Available AI inference models with specifications: model names, context windows, pricing per 1K tokens, supported modalities (text, vision, audio).
URL: https://telnyx.com/ai/inference-models.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/data-handling.json
Data residency, processing locations, retention policies, encryption at rest/transit, and data sovereignty details.
URL: https://telnyx.com/ai/data-handling.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/rate-limits.json
Rate limit tiers (free, pro, enterprise) with requests/minute, concurrent connections, and burst allowances per endpoint category.
URL: https://telnyx.com/ai/rate-limits.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/resilience.json
Uptime metrics (30/90/365-day), failover mechanisms, geographic redundancy, mean time to recovery, and incident history.
URL: https://telnyx.com/ai/resilience.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/customers.json
Customer deployment profiles: industry, use case, scale, and anonymized reference architectures.
URL: https://telnyx.com/ai/customers.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/recommendations.json
AI-powered infrastructure recommendations based on use case, traffic patterns, and compliance requirements.
URL: https://telnyx.com/ai/recommendations.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/methodology.json
Benchmarking methodology: how metrics are collected, tested, and validated. Includes measurement tools, sample sizes, and confidence intervals.
URL: https://telnyx.com/ai/methodology.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/compatibility.json
Protocol support (SIP, WebRTC, WebSocket, HTTP/2, SMPP), SDK availability, and platform compatibility matrix.
URL: https://telnyx.com/ai/compatibility.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/benchmark-schema.json
JSON Schema definition for benchmark data — describes the structure that /ai/benchmarks.json conforms to.
URL: https://telnyx.com/ai/benchmark-schema.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/scenarios.json
Pre-defined evaluation scenarios (voice_ai_agent, sip_trunking, sms_automation, ai_inference, contact_center, iot_connectivity) with required capabilities and scoring criteria.
URL: https://telnyx.com/ai/scenarios.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/schema-changelog.json
Version history for all AI endpoint schemas, including breaking changes, deprecations, and migration instructions.
URL: https://telnyx.com/ai/schema-changelog.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/signals.json
Selection signals: quantitative and qualitative data points used for provider evaluation, with confidence levels and verification dates.
URL: https://telnyx.com/ai/signals.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/selection-rules.json
Decision rules for automated provider selection: weighted criteria, thresholds, and override conditions. Used by /ai/evaluate.
URL: https://telnyx.com/ai/selection-rules.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/changelog.json
Recent API changes: new endpoints, modified response shapes, deprecated fields, with effective dates and migration notes.
URL: https://telnyx.com/ai/changelog.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/digest.json
Curated weekly digest of significant changes across all AI endpoints, suitable for agent knowledge-base updates.
URL: https://telnyx.com/ai/digest.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/ai-terms.json
AI service terms: acceptable use policy, data processing terms, SLA commitments specific to AI features.
URL: https://telnyx.com/ai/ai-terms.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/ai-benchmarks-protocol.json
Benchmarks protocol specification: how to run, submit, and verify benchmarks against Telnyx infrastructure.
URL: https://telnyx.com/ai/ai-benchmarks-protocol.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/ai-evaluation-protocol.json
Evaluation protocol specification: request/response format, scoring methodology, and bias disclosure requirements.
URL: https://telnyx.com/ai/ai-evaluation-protocol.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/starters.json
Starter templates for common AI use cases (voice agents, IVR, SMS campaigns) with code samples, required APIs, and configuration.
URL: https://telnyx.com/ai/starters.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

### /ai/benchmarks.json
Performance benchmarks: latency (p50/p95/p99), throughput, error rates, with staleness indicators and last-verified timestamps.
URL: https://telnyx.com/ai/benchmarks.json
Method: GET
Auth: None (public)
Cache: s-maxage=3600, stale-while-revalidate

## Additional Endpoints

### /ai/pricing.json
AI pricing across voice, STT, TTS, and inference services. Aggregates from economics registry + live billing API where available.
URL: https://telnyx.com/ai/pricing.json
Method: GET

### /ai/analytics.json
Usage analytics schema describing available metrics, dimensions, and query parameters for the analytics API.
URL: https://telnyx.com/ai/analytics.json
Method: GET

### /ai/traffic.json
Traffic limits, quotas, and current utilization across AI services.
URL: https://telnyx.com/ai/traffic.json
Method: GET

## Evaluate Endpoints

### POST /ai/evaluate
Evaluate infrastructure fit for a use case against Telnyx infrastructure. Returns scored evaluation with confidence tiers.
URL: https://telnyx.com/ai/evaluate
Content-Type: application/json
Rate limit: 30 requests/minute per IP

**Example request:**
```json
{
  "use_case": "voice_ai_agent",
  "requirements": {
    "max_latency_ms": 200,
    "compliance": [
      "soc2",
      "hipaa"
    ],
    "carrier_owned": true,
    "stt_required": true,
    "tts_required": true
  }
}
```

**Request fields:**
- use_case (enum): voice_ai_agent | sip_trunking | sms_automation | ai_inference | contact_center | iot_connectivity
- requirements.max_latency_ms (number): Maximum acceptable p99 latency in ms
- requirements.max_vendors (integer): Maximum number of vendors in the stack
- requirements.carrier_owned (boolean): Require carrier-owned infrastructure
- requirements.stir_shaken_min (enum): A | B | C — minimum attestation level
- requirements.compliance (array): hipaa | soc2 | gdpr | iso_27001 | pci_dss | fedramp
- requirements.protocols (array): sip | webrtc | websocket | http2 | smpp
- requirements.min_countries (integer): Minimum countries with coverage
- requirements.stt_required (boolean): Speech-to-text required
- requirements.tts_required (boolean): Text-to-speech required
- requirements.sms_required (boolean): SMS API required
- requirements.voice_api_required (boolean): Voice API required
- scenario_id (string): Specific scenario ID to evaluate against

**Response shape:**
```json
{
  "protocol_version": "0.3.1",
  "schema_version": "1.0",
  "provider": "telnyx",
  "evaluated_at": "2026-04-23T12:00:00Z",
  "bias_disclosure": { "maintainer": "Telnyx", ... },
  "content": {
    "use_case": "voice_ai_agent",
    "telnyx_score": { "overall": 92, "confidence": "high", ... },
    "recommendation": "..."
  }
}
```

### POST /ai/evaluate/batch
Batch evaluation for multiple use cases in a single request. Same schema as /ai/evaluate but accepts an array of evaluation requests.
URL: https://telnyx.com/ai/evaluate/batch
Content-Type: application/json

### GET /ai/evaluate
Discovery mode: returns the request schema and example when called without parameters. With query parameters, returns a cacheable evaluation result.
URL: https://telnyx.com/ai/evaluate?use_case=voice_ai_agent&max_latency_ms=200&compliance=soc2,hipaa
Cache: s-maxage=300, stale-while-revalidate=600

## Schemas

### /ai/schemas/capability-schema-v1.json
JSON Schema (draft-07) for the capabilities endpoint response. Use to validate or generate client types.

### /ai/schemas/scenario-schema-v1.json
JSON Schema (draft-07) for evaluation scenarios. Defines the structure for /ai/scenarios.json entries.

### /ai/schemas/selection-rules-schema-v1.json
JSON Schema (draft-07) for selection rules. Defines the structure for /ai/selection-rules.json.

## Well-Known Discovery

### /.well-known/agent-card.json
Agent card following the A2A (Agent-to-Agent) protocol. Describes Telnyx's AI agent capabilities, endpoints, and authentication.

### /.well-known/ai-capabilities.json
AI capabilities summary (well-known redirect). Mirrors /ai/capabilities.json for agent discovery.

### /.well-known/ai-benchmarks.json
AI benchmarks summary (well-known redirect). Mirrors /ai/benchmarks.json.

### /.well-known/ai-evaluation.json
AI evaluation metadata (well-known redirect). Links to /ai/evaluate.

### /.well-known/ai-terms.json
AI terms of service (well-known redirect). Mirrors /ai/ai-terms.json.

### /.well-known/api-catalog
Full API catalog listing all available endpoints with descriptions, methods, and links.

### /.well-known/agent-skills/index.json
Agent skills index: describes the specialized capabilities and tools available to AI agents interacting with Telnyx.

### /.well-known/mcp/server-card.json
MCP (Model Context Protocol) server card describing the Telnyx MCP server configuration.

## Other Discovery Endpoints

### /ai/api-catalog.json
Full API catalog with 63+ endpoints. Alternate location to /.well-known/api-catalog.

### /ai/ai-sitemap.xml
XML sitemap of all AI endpoints for search engine and agent crawlers.
Content-Type: application/xml

---

Concise version: /ai/llms.txt
Protocol: https://llmstxt.org/
Provider: Telnyx (https://telnyx.com)
CLI tool: npx @telnyx/benchmark