EnsembleData pricing and an alternative

EnsembleData sells monthly subscriptions. Each plan gives you a number of units per day, and each call uses units. Unused units vanish at 00:00 UTC. The free tier gives 50 units a day, no card needed.

• —Entry plan. Wood, $100 a month for 1,500 units a day.
• —Top listed plan. Platinum, $1,400 a month for 50,000 units a day. Diamond is a custom quote.
• —Units per call. It depends on the endpoint. Basic user info costs 1 unit. A rich profile with posts costs 10. Some endpoints charge per result returned.
• —Failed calls. Only internal errors go uncharged.
• —Rate limits. No limit is published.

Example: the $100 plan gives about 45,000 units in a 30-day month. If every call costs 1 unit and you use the full quota every day, that is about $2.22 per 1,000 calls. At 10 units a call, it is 150 calls a day. Units you do not use on a quiet day are lost.

• —Midnight UTC. Unused units are gone at the end of the day. A quiet Sunday is paid for and thrown away.
• —Unit weights. A basic call is one unit, a rich profile with posts is ten. The plan you need depends on which endpoints you call, not how many.
• —Subscription first. The bill is the same whether you used the quota or not.
• —Bursts. A settlement run or a launch day needs more than a daily quota allows, and the next plan up is a big step.

• —Steady daily loads. If you use the same amount every day, a fixed plan can come out cheaper than per-request pricing.
• —Breadth. YouTube, Twitch, Snapchat, Threads, if you need them today. We go deep on Instagram, TikTok, X, Reddit, and YouTube, and add platforms as customers ask.

• —No quota. Use 50 requests one day and 50,000 the next. You pay for what you used, after the fact.
• —One call, one price. A profile costs the same as a post or a page of comments. No unit weights.
• —Free when you cache. A 24-hour cache hit costs $0.0005. A cache miss uses the live price, and a 30-day hit is free.
• —Hard limits. Your balance is the limit. You can also set a monthly request cap on each key.

1. 01Start with a Test key. Create a Test key in the dashboard. Use the synthetic identifiers in the quickstart to validate your integration without charges.
2. 02Map your calls. Match your required data to the API reference. Check endpoint coverage, parameters, and returned fields before switching. Keep unsupported operations with your current provider. The table below lists common Openhandle calls.
3. 03Update paths, parameters, and authentication. Use https://api.openhandle.dev/v1 with Authorization: Bearer followed by your key. Update each endpoint path and parameter using the API reference. Profile usernames need an @ prefix in REST paths. Use the documented identifier for each post.
4. 04Map response fields. Successful responses contain platform, resource, capturedAt, source, and data. Map shared fields and platform-specific fields using the response schemas. Keep unavailable metrics as null.
5. 05Handle pagination and errors. Pass meta.cursors.next back as cursor for list requests. Cursors from another provider cannot be reused. Handle Openhandle error codes and retry only when retryable is true. See pagination and errors.
6. 06Set a cap. Pick a monthly spending cap before you go live. We stop before it, so a bad loop cannot run up a bill.
7. 07Validate before switching. Check success, empty, private, and not-found cases with a Test key. Then move to Live and compare a small sample with your current provider before switching production traffic.