As for J-10CE, the rumored, and I emphasize rumored, export cost is 40mn, presumably not including spares, support, and maintenance. The planned JF-17 Block III cost is 32 million. The comparable Western system is the F-16V, which has a flyaway cost of about 60-70mn. We will have to wait until the J-10CE is exported to have an affirmative notion of its production cost.
==
Put another way, AFB doesn't believe that the J-20 can 2:1 F-35s. My point isn't whether the J-20 can or cannot, because it depends not only on airframes, but on supporting elements (EODAS B-21 / H-20 support and counterstealth AEW&C) and subsystems (interception XVRAAMs like the PL-21 / PL-15 and AIM-260, laser point defense on future F-35s, micromissiles and active chaff on the F-35), but whether the J-20 NEEDS to do so.
What people arguing for lower prices (China's cheap, labor is cheap [but getting more expensive!]) basically want to say is that the J-20 shouldn't need to 2:1 F-35s. Which gives you a huge problem, because China is going to be facing 1200 F-35s (60% tilt of total production into East Asia theater), implying that China has to build 1200 J-20s to counter.
Moreover, on a symbolic level, managing to 1:1 or resorting to 1:1-ing a lightweight-equivalent / middleweight fighter just says your technology is garbage. It's as if you couldn't take out T-34s on a 1:1 ratio except with Tigers.
We're no longer in a People's War era wherein China is committing to taking 10:1 casualty ratios to deter an enemy, the overall game is a technological race. Getting 2:1 with heavyweight vs lightweight is the ideal circumstance, and getting 4:1 or 8:1 by having a generational advantage is also the ideal circumstance.