Fitting $f(Q,T)$ gravity models with a $\Lambda$CDM limit using H(z) and Pantheon data

We proposed five $f(Q,T)$ models, which are an extension of symmetric teleparallel gravity, where $Q$ is the non-metricity and $T$ is the trace of the stress-energy tensor. By taking specific values of their parameters, these models have a $\Lambda$CDM limit. Using cosmic chronometers and supernovae Ia data, we found that our models are consistent with $\Lambda$CDM at a 95\% confidence level. To see whether one of these models can challenge $\Lambda$CDM at a background perspective, we computed the Bayesian evidence for them and $\Lambda$CDM. According to it, the concordance model is preferred over four of them, showing a weak preference against $f(Q,T) = -Q/G_N + bT$ and $f(Q,T) = -(Q+2\Lambda)/G_N +bT$, a substantial preference against $f(Q,T) = -(Q+2 H_0^2 c (Q/(6H_0^2))^{n+1})/G_N + bT$, and a strong preference against $f(Q,T) = -(Q+2H_0^2c(Q/(6H_0^2))^{n+1} + 2\Lambda)/G_N + bT$. Interestingly, a model includying a $T^2$ dependence ($f(Q,T) = -(Q+2\Lambda)/G_N - ((16\pi)^2 G_N b)/(120 H_0^2) T^2$) showed a {substantial} preference against $\Lambda$CDM. {Therefore, we encourage further analyses of this model to test its viability outside the background perspective.}