There are several companies currently launching experimental satellites and it seems that there will soon be ways to get internet access either directly or via a small ground station that relays via local WiFi.
If the satellites are in very low earth orbit, some as low as 200km, the latency can be comparable to ground based internet services.
SpaceX was granted permission to launch 7,000 satellites for Starlink internet provision back in 2018. The full plan involves 30,000 satellites. The plan is offer internet to every part of the earth by having at least one satellite visible at any time. Initial service might start as early as 2020.
There's a rather out of date FAQ here on Reddit.
Amazon's Project Kuiper involves 3,236 satellites at various low earth orbit heights. "784 satellites at 367 miles, 1,296 satellites at 379 miles, and 1,156 satellites at 391 miles". Rather than covering the entire globe, they're going for a band north and south of the equator that covers 95% of the population.
OneWeb provides internet for business jets. Their system is made up of both a swarm of satellites and ground stations.
OneWeb recently completed tests from Seoul, South Korea using just six satellites and showed they could deliver "high-speed, low-latency services at speeds of more than 400 Mbps, enabling the fastest real-time video streaming in Full HD". Latency is reported to be 40 milliseconds but with an average of just 32ms.
Swarm has been granted permission for a constellation of 150 LEO satellites for provision of non-voice Mobile-Satellite Services (MSS). While targeting IoT customers, which suggests low bandwidth. Reportedly 1kbps initially, rising to 2.7kbps. Not much good for internet but useful for tracking and messaging.
New internet competition?
The big question with all of these new options is the price. If players choose to go after the elite end of the market, such as business jets, then it won't affect many of us. If players, as Amazon is reported to be doing, target advanced markets like the US, then this could threaten terrestrial providers.
My hope is that we get services that target under served markets such as outback Australia and the Pacific Islands. Particularly in developing countries prices will need to be low to be affordable.
The equipment needed to talk to low earth orbit satellites must have rapidly steerable directional antennas but just as GPS receivers have become small and cheap, this will also become affordable at scale.
Can satellite internet be blocked?
Some countries filter the internet. Australia is on this list as a surveillance state. Unlike bans on receiving satellite TV (which used to require large dishes that could be seen), it's going to be hard to detect users of satellite internet. Like jamming of GPS, presumably these satellites or the receivers could be overwhelmed by strong RF until they comply with local requirements.
Please let me know in the comments if there are other players I've missed.