Astrology API reviews.
Builders ship faster with Roxy.

Cloned the open source AI chatbot starter, added our API key, deployed to Vercel. Working multi-domain spiritual advisor in production the same day. Astrology, tarot, and numerology all through one conversation.

The llms.txt and MCP support means AI agents discover Roxy capabilities on their own. We built a Gemini agent that uses astrology, tarot, and dream interpretation tools without hardcoding a single endpoint path.

Built a multi-agent system where one agent handles birth chart analysis and another does tarot interpretations. The MCP server lets each agent discover only the tools it needs. Clean separation of concerns.

I am not a developer and I got daily horoscopes working in my app using just the interactive API reference to test calls. Copy-pasted the curl examples into my workflow. Whole thing took an hour.

The Expo starter app got us from zero to a working tarot reading app on TestFlight in one afternoon. Three-card spread, daily draws, yes/no oracle. All the endpoints just work, no configuration.

Told Cursor to build me a kundli generation app using Roxy. It found the MCP server, connected to it, and had a working prototype with birth charts and dashas in 20 minutes. I barely wrote any code myself.

MCP server connected to Claude Desktop in two minutes.

Cursor picked up all the tools without any config file.

Windsurf detected the MCP endpoint and started using the right tools right away.

Switched our agent stack to one Roxy MCP. Tool calls per agent dropped noticeably and routing got cleaner.

Gemini agents discover the tarot tools through llms.txt with zero hardcoding.

Custom GPT loaded the spec and started running birth charts inside the same conversation.

Wired the MCP server into our internal team bot. We use it for weekly horoscopes now.

I was skeptical one MCP would handle tool selection across all 18 domains. It does.

Our agent picks Roxy tools correctly without any prompt engineering.

Wired Roxy MCP into our internal agent stack. Tool routing for birth chart and tarot works without per-call prompt tuning.

Built a daily horoscope WhatsApp bot in a weekend using the API.

Vedic dasha periods matched my reference desktop software within a day across the charts I tested.

Expo starter compiled and ran on first try. Zero dependency warnings.

I had been a customer of two other astro APIs. Roxy replaced both and the KP cusps line up with my reference.

Panchang endpoint returned Tithi and Nakshatra exactly matching DrikPanchang for the dates I checked.

Subscribed Friday, shipped the MVP Sunday night. The API just works.

Tarot draw with the seed parameter was the missing piece for our reproducible reading feature.

Numerology life path returns master numbers correctly. We had been hand-rolling that.

I-Ching hexagram lookup with changing lines was perfect for our divination chatbot.

Daily horoscope content has unique copy per sign. Other providers serve identical text across signs.

INR pricing flow felt local from day one. No surprise foreign card fees on my Indian card.

Set up a Vedic kundli widget for my blog in two hours. Subscribers seem to spend more time on the page.

Used the AI chatbot starter as a teaching example. Students built their own forks the same day.

Latencies have been steady under our load. p99 sits comfortably below our budget.

The crystal recommendations endpoint was the surprise hit of our app. Users keep asking for more.

Angel number lookup integrated cleanly into our notification engine. The localized meanings were a nice plus.

Dream symbol API filled the last gap in our journaling feature.

Composite chart endpoint covers both midpoint and time-mean methods, which most providers do not.

Synastry API gave us all the inter-aspects we needed for the dating UI.

Birth chart with Whole Sign houses returns identical placements to my desktop reference software.

Solar return endpoint returns the correct ascendant for the return chart, finally.

Lunar return calculation timing matched my reference data closely on the dates I checked.

First-time API user. The interactive docs let me test every endpoint before I wrote any code.

OpenAPI spec is so clean we generated SDKs for two languages we did not even plan to support.

I demo Roxy at meetups now. The MCP plus chatbot starter is the cleanest live demo I have.

Migrated my team to Roxy from another provider. Onboarding took less than a day.

Birth chart in our React Native app shipped in two days. The Expo starter saved us a week of plumbing.

API key activation was instant after checkout. No support ticket needed.

Rate limits are honest. The plan caps match what we measured under load.

Aspect calculations let us configure orbs per request. Most providers fix the orb at three degrees.

I cross-checked a stack of charts against my own reference software. Roxy was within tolerance every time.

Daily yes-no tarot endpoint became our most-used feature.

Three card spread is the most engaging surface on our landing page.

Numerology endpoint preserves master numbers 11, 22, 33. Most calculators round them down.

Subscribed for Vedic, stayed for the breadth. We use seven of the 18 domains in production now.

Time zone handling for birth chart is correct across DST boundaries. We had bugs with our previous provider.

No need to self-host the MCP. The hosted endpoint holds up under our concurrency.

The OpenAPI examples auto-fill our test fixtures. We deleted half our seed data.

Cloudflare Workers integration was a one-liner. No SDK rewrite required.

Hindi locale on tarot is properly localized, not machine-translated text.

Spanish locale on horoscopes lets us serve our LATAM users in their language.

Vedic locale in Hindi script was the deal-closer for our family-tier customers.

Hard to overstate how much time the documentation saved. I never had to ask anyone anything.

Vimshottari Mahadasha returns full sub-period structure. Antar dasha boundaries match my reference.

Roxy was the only provider we evaluated that hit our latency budget at our concurrency.

MCP server returned the right tool combination for our birth chart conversational UX. No tool routing bugs.

The onboarding email pointed me to the right starter for my use case. Skipped a week of research.

We use the moon phase data in our farming-related app. Lunar timing is accurate to the hour.

Daily transits with house overlay is the cleanest content layer I have seen for horoscope apps.

API key rotation worked without downtime when we cycled credentials in CI. No surprises.

API itself is solid. A small free tier for prototyping before the first paid month would lower the entry bar.

Numerology endpoint is great. Documentation could include a quick-start for the master number rule.

OpenAPI spec is well structured. A first-party Postman collection would shave the first-day setup time.

Vedic API works correctly for our use case. The panchang explainer text could be more verbose for end users.

The MCP server documentation could be clearer for first-time users. Once we got past the initial setup it has been a non-issue.

Solid Expo starter. A richer demo screen out of the box would help non-developers understand the value faster.

A streaming endpoint for long horoscope generations would help our chatbot UX. JSON-only is fine for now.

Clean API surface. Add a CSV export option for the dasha tables and it is a five for me.

Works as advertised. The pricing page could explain currency selection more clearly during checkout.

API itself is fine. Onboarding could include a checklist of recommended endpoints by use case.

A public status page with historical p95 numbers would help SLA conversations on our side.

Functional and well documented. The angel number examples could include context for less common numbers.

Would love an embeddable widget script for daily horoscope without a backend integration. Hand-rolled it for now.

A few starter app screenshots in docs do not match the current default theme. Minor, but slowed me down briefly.

Spec is well typed. The error response code field took a minute to grep the docs to find.

Could use clearer guidance on when to use natal versus transit endpoints. Eventually figured it out from the OpenAPI spec.

Calculations are stable and accurate. The KP significators endpoint could include a worked example in the docs.

Cache headers on the daily horoscope endpoint took a minute to figure out. Documenting the strategy per endpoint would help.

API is solid. The Python SDK could expose async helpers in a future release.

API works correctly. The MCP per-product URL pattern takes a beat to grok at first.

Infrastructure has held up under our load. An IP allowlist option for keys would help our compliance review.

A 7-day starter discount instead of full price month one would have closed the deal sooner. Glad I tried it anyway.

The transit aspects endpoint result set is large. Pagination or a default minor-aspect filter would help.

Composite chart docs could explain the difference between the methods more concretely. The endpoint itself is fine.

Great API surface overall. The X-API-Key header docs could surface earlier in the quickstart for first-time integrators.

The MCP per-tool descriptions could be more terse. Token cost adds up across longer agent sessions.

A first-class JS embed for sign-of-the-day cards would be a nice add. The API is otherwise straightforward.

Wish the SDK retried automatically on 429 with backoff. Wrote it ourselves and moved on.

Would love a birth time rectification helper as a future feature. The natal chart accuracy is already there.

Tarot deck variant support was not obvious from the docs at first. Worked it out by reading the OpenAPI spec.

Pricing in INR was a relief. UPI as a checkout option would be even better for our small recurring spend.

Integration was smooth. An iOS shortcut sample for the API would help non-pro mobile users get going.

Turkish localization is well done. Arabic would be a nice add down the line.

The biorhythm endpoint output could include a chart-friendly time series by default. Works fine, just an extra reshape on our end.

API has held up under our scheduled job load. A documented best-practice for retry-with-backoff on 429 would help SDK consumers.

The Vedic yogas list is huge. A filter by category would make it easier to navigate from a UI.

The chat starter is React-only. A Vue or Svelte port would broaden adoption beyond the React crowd.

Concurrent request handling at the SDK level would smooth our background job worker. Otherwise the API is fine.

Audit logs for API key activity in the dashboard would be nice for our compliance review.

Works well with our agents. Streaming on the longer horoscope endpoints would help the UX though.

The 404 error body could include a short list of likely causes. Otherwise the spec types are clean to consume.

Wish the Vedic Lahiri ayanamsha allowed user-supplied values for non-standard schools. Default Lahiri is correct, just inflexible for niche schools.

Caching guidance per endpoint would help us tune our edge config. Otherwise the API has been steady for our workload.

An auto-detected timezone helper at the SDK level would reduce setup friction for mobile flows.

The starter README assumes some Vedic vocabulary. A short glossary upfront would help engineers ramp faster.

Decent. The free tools page is useful but I had to check three pages before realizing the API itself is paid only.

It works. The onboarding email had a few different starter recommendations and I was unsure which one fit my use case.

API is fine. The pricing tier names could be more transparent about request limits without clicking into details.

Documentation has a learning curve if you do not already know astrology jargon. A short onboarding glossary would help.

Works once dialed in. I had timezone confusion at first. A note in the docs about local time plus offset would have saved me time.

MCP latency on cold start was higher than I expected for streaming chat use cases. Steady on warm connections.

It does what it says. End-to-end test examples for the API in the starters would be a nice add.