For a while it has looked as if the C7 v2 was a very good choice if one wanted to run open firmware. In particular, it uses an Atheros radio that has open drivers. Based on my criteria, OpenWrt is the firmware to run (or something based on it). DD-WRT and others have problems that make them not open by my standards. I haven't actually run OpenWrt on a C7. But I learned a few things last night that dismayed me. You can find these in "Note:" entries on the wiki page but they don't stick out. <https://wiki.openwrt.org/toh/tp-link/archer-c5-c7-wdr7500> 0) It is "OpenWrt", not "OpenWRT". 1) Recent stock firmware from TP-Link locks out flashing third-party firmware. <http://ml.ninux.org/pipermail/battlemesh/2016-February/004379.html> If you have a C7 with older stock firmware, don't upgrade it to the current stock firmware if you ever hope to use OpenWrt. There are apparently work-arounds but they seem a bit intricate. 2) The C7 (and a number of other routers) have a hardware fast-path to do NAT. This hardware is undocumented and hacky so OpenWrt will never use it. Without using it it is impossible to get gigabit wire-speed NATting. But the stock firmware does use it. <https://dev.openwrt.org/ticket/11779> 3) For mysterious reasons OpenWrt is significantly less efficient handling 5GHz 802.11ac. You won't notice this unless you are using a 3-antenna client (or, I speculate, multiple 2-antenna clients) <https://forum.openwrt.org/viewtopic.php?id=53703> This thread is pretty frustrating because the first message very carefully posts observations and yet most of the following 168 messages ask questions already answered or veer off-topic. There appears to be no resolution. All this reinforces my decision to build gateways out of small PCs rather than consumer routers. I do use an off-the-shelf wireless router as an access point.