In this video here Warren Mosler states that all economic models fail to predict inflation ("All their models show it's indeterminate"). I am looking for some references to support that claim... Or is Warren plain wrong?
Depends on what the author meant by:
All their models show its indeterminate.
If you would want to take this in a literal sense then no. An easy counter example would be some bachelor level monetary model or Philipps curve.
However, there is a famous indeterminacy problem in macroeconomics which is real in many models. The problem is that in many (but not all) macroeconomic models forward looking monetary interest rates setting rules such as Taylor rule with finite horizon may lead to real indeterminacy which means that when you shock your system there might be many different paths that will make real variables return to the long run equilibrium, and consequently also to indeterminacy in the rate of inflation.
But there are also mainstream models outside MMT that solve this problem. For example, even in this paper from ECB authors show that just change in interest rate setting rule can resolve this issue. Moreover, again this is definitely not issue in “All economic models”.
However, if the author was exaggerating and by that meant that there is a serious indeterminacy issue in many models based on simple finite horizon interest rate setting rules, which is still an active research topic in mainstream macroeconomics then in that sense the statement would be accurate.