You have a few options for this:
If you can pay for this, use commercial API monitoring solutions like Runscope or APImetrics. They're easy to setup, have no maintenance overheads, and show you detailed data and alerting etc. The problem with this approach is that they cost money, and if you have a miniscule ops budget or this is a hobby project, it can be hard to justify spending on this.
...which brings me to the alternative: roll your own. If you have your own system monitoring and metrics setup (Nagios/Icinga/Sensu, and Grafana etc), you can write check scripts for whatever data you need (availability, response time or latency etc). Some of these tools integrate with time series databases like graphite/influxDB, so you can push the check data to that and use a visualization tool like Grafana to show graphs and charts based on this data (here's a blog post describing Icinga-Grafana integration)