makesplat

Quickstart

From API key to downloadable .ply in four HTTP calls. Any language, any stack — if it can speak HTTPS, it can make splats.

1 · Get an API key

Sign up, open the dashboard, and create a key. It looks like sk_live_... — store it like a password, it's shown only once. Send it on every request:

Authorization: Bearer sk_live_your_key_here

2 · Reserve a scan

This returns a job_id and a presigned upload_url (valid 1 hour). Nothing is charged yet.

curl -X POST https://api.makesplat.com/v1/scans \
  -H "Authorization: Bearer sk_live_..."

# → { "job_id": "...", "upload_url": "https://...", "expires_in_seconds": 3600 }

3 · Upload your video

PUT the file straight to the upload URL — no auth header needed, the URL is signed. MP4/MOV up to 500 MB. Orbit the object slowly with good lighting for best results.

curl -X PUT --upload-file dish.mp4 "<upload_url>"

4 · Start processing

Credits are charged here (1.0 for defaults). Pass parameters to trade cost for quality — or call /v1/scans/estimate first to see the price of any combination.

curl -X POST https://api.makesplat.com/v1/scans/<job_id>/start \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{"params": {"steps": 15000, "downscale": 2, "matching": "sequential"}}'

# → { "job_id": "...", "status": "queued" }
# 402 = not enough credits. Failed jobs refund automatically.

5 · Poll until done, download

Processing takes ~10-40 minutes depending on queue and quality. Poll every 15-30 s; when status isdone, the response carries a signed ply_url (valid 24 h).

curl https://api.makesplat.com/v1/scans/<job_id> \
  -H "Authorization: Bearer sk_live_..."

# → { "status": "done", "ply_url": "https://...", "ply_size_mb": 64.2,
#     "elapsed_sec": 1240, "credits_charged": "1.0", ... }

curl -o dish.ply "<ply_url>"

Parameters

ParamRangeDefaultEffect
steps1,000-30,00015,000Training iterations — detail vs cost
downscale1, 2, 421 = full resolution (pricier), 4 = quarter
num_frames30-300120Frames extracted from your video
matchingsequential · vocab_tree · exhaustivesequentialCamera-pose matching strategy

Batch processing

There's no special batch endpoint — just loop. Reserve and start N jobs, then poll them together. Jobs run in parallel on our GPU fleet.

import requests

API = "https://api.makesplat.com"
H = {"Authorization": "Bearer sk_live_..."}

jobs = []
for video in ["a.mp4", "b.mp4", "c.mp4"]:
    r = requests.post(f"{API}/v1/scans", headers=H).json()
    requests.put(r["upload_url"], data=open(video, "rb"))
    requests.post(f"{API}/v1/scans/{r['job_id']}/start", headers=H, json={})
    jobs.append(r["job_id"])

# poll until all done, then download each ply_url

Rate limits & rules