diff --git a/docs/api/modeling.rst b/docs/api/modeling.rst
index ba317bd0..ab4da4b4 100644
--- a/docs/api/modeling.rst
+++ b/docs/api/modeling.rst
@@ -35,23 +35,13 @@ Input into an ``Analysis`` class to customize the behaviour of a CTI model-fit p
 Non-linear Searches
 -------------------
 
-A non-linear search is an algorithm which fits a model to data.
-
-**PyAutoCTI** currently supports three types of non-linear search algorithms: nested samplers,
-Markov Chain Monte Carlo (MCMC) and optimizers.
+A non-linear search is an algorithm which fits a model to data. The searches are provided by
+**PyAutoFit**, and every one can be used with **PyAutoCTI**: see PyAutoFit's
+:ref:`capability matrix <autofit:search_capability_matrix>`, which lists every search, what it
+supports (JAX, gradients, evidence, posterior kind, install) and links to its API reference.
 
 .. currentmodule:: autofit
 
-.. autosummary::
-   :toctree: _autosummary
-   :template: custom-class-template.rst
-   :recursive:
-
-   Nautilus
-   DynestyDynamic
-   Emcee
-   LBFGS
-
 Priors
 ------
 
diff --git a/docs/conf.py b/docs/conf.py
index a6398525..855586f9 100644
--- a/docs/conf.py
+++ b/docs/conf.py
@@ -88,6 +88,9 @@ def clone_repo(name: str, url: str):
 
 intersphinx_mapping = {
     "python": ("https://docs.python.org/3", None),
+    # PyAutoFit owns the non-linear searches; api/modeling.rst links to its
+    # generated capability matrix (search_capability_matrix).
+    "autofit": ("https://pyautofit.readthedocs.io/en/latest", None),
     "sphinx": ("https://www.sphinx-doc.org/en/main", None),
 }