| Compute Engine | ✓VedIntel™ Precision Engine — 100% local | ✗External API dependency |
| Third-party downtime risk | ✓Zero — runs on our servers | ✗Yes — upstream outages affect your app |
| AI Interpretation Layer | ✓Claude AI — Professional and Enterprise plans | ✗Not available on most providers |
| Permanent Natal Cache | ✓SHA-256 cache — never recomputed | ✗No caching guarantee — re-billed each call |
| Free Plan | ✓500 calls, 7-day trial, 128 of 135 endpoints | ✗50 calls, limited endpoint set |
| Starter Plan | ✓$49/mo — 100K calls, 129 endpoints | ✗$49–$150/mo for fewer calls |
| Unified Endpoint Access | ✓Same 129 endpoints on every paid plan (AI narratives are Professional+) | ✗Endpoint categories sold separately |
| Response Envelope | ✓status + response + remaining_api_calls | ✗status + response only |
| Ayanamsa options | ✓12 (Lahiri default, Raman, KP, Pushya Paksha + 8 more) | ✗Lahiri only on most |
| Dasha systems | ✓Vimshottari, Yogini, Char Dasha | ✗Varies — typically Vimshottari only |
| PDF Reports | ✓AI-written narrative per report, not templated | ✗PDF on higher plans only |
| Languages supported | ✓23 languages | ✗Varies (usually 9–12) |
| MCP Server | Live — npm @vedintelastroapi/mcp-server | Varies |
| WordPress Plugin | Live on WordPress.org | Varies |
| Rate limit (base) | 20 req/sec | 10–20 req/sec |
| Rate limit (enterprise) | 500 req/sec | Custom |