All high field superconductors producing magnetic fields above 12 T are brittle. Nevertheless, they will probably be the materials of choice in commercial tokamaks because the fusion power density in a tokamak scales as the fourth power of magnetic field. Here we propose using robust, ductile superconductors during the reactor commissioning phase in order to avoid brittle magnet failure while operational safety margins are being established. Here we use the PROCESS systems code to inform development strategy and to provide detailed capital-cost-minimised tokamak power plant designs. We propose building a 'demonstrator' tokamak with an electric power output of 100 MW e , a plasma fusion gain Q plasma = 17, a net gain Q net = 1.3, a cost of electricity (COE) of $ 1148 (2021 US) per MW h (at 75 % availability) and high temperature superconducting operational TF magnets producing 5.4 T on-axis and 12.5 T peak-field. It uses Nb-Ti training magnets and will cost about $ 9.75 Bn (2021 US). An equivalent 500 MW e plant has a COE of $ 608 per MW suggesting that large tokamaks may eventually dominate the commercial market. We consider a range of designs optimised for capital cost (as the reactors considered are pilot plants) consisting of both 100 MW e and 500 MW e plants with each of two approaches for the magnets: training and upgrading. With training magnets, the plant is cost-optimised for REBCO TF magnets. For a 100 MW e plant, the Nb-Ti training magnets typically produce 70 % peak field on the toroidal field coils compared to REBCO magnets, 65 % peak field on the central solenoid and cost ≈ 10 % of the total machine cost. Training magnets could in principle be reused for each of say 10 subsequent (commercial) machines and hence at 1 % bring only marginal additional cost. With upgrade magnets the plant is more expensive -first it is cost-optimised for Nb-Ti and then upgraded to REBCO coils. The upgrade increases the net electrical output from 100 to 280 MW e with an ≈ 25 % increase in reactor capital cost. We also evaluate likely advances in fusion technology and find that technologies on the horizon will probably not bring further large reductions in capital cost, and that REBCO magnets are generally stress-limited rather than current density limited. We conclude that: the fusion community should develop high B c2 alloys specifically for fusion applications; superconductors should be tested under operational-like radiation at cryogenic temperatures; and that we should proceed now with detailed design and construction of a prototype fusion power plant that integrates and de-risks all the key technologies including high temperature superconducting cables and joints using remountable training magnets, and hence is the last tokamak before commercialisation of fusion energy.