diff --git a/.github/workflows/test_benchmark_collection_models.yml b/.github/workflows/test_benchmark_collection_models.yml
index 7c7aacb992..7b1568b1fc 100644
--- a/.github/workflows/test_benchmark_collection_models.yml
+++ b/.github/workflows/test_benchmark_collection_models.yml
@@ -61,6 +61,7 @@ jobs:
       run: |
         git clone --depth 1 https://github.com/benchmarking-initiative/Benchmark-Models-PEtab.git \
           && export BENCHMARK_COLLECTION="$(pwd)/Benchmark-Models-PEtab/Benchmark-Models/" \
+          && pip3 install -e $BENCHMARK_COLLECTION/../src/python \
           && AMICI_PARALLEL_COMPILE="" tests/benchmark-models/test_benchmark_collection.sh
 
     # run gradient checks
diff --git a/tests/benchmark-models/test_benchmark_collection.sh b/tests/benchmark-models/test_benchmark_collection.sh
index f8ccd0c1cd..51f6ad06d5 100755
--- a/tests/benchmark-models/test_benchmark_collection.sh
+++ b/tests/benchmark-models/test_benchmark_collection.sh
@@ -87,6 +87,12 @@ script_path=$(cd "$script_path" && pwd)
 
 for model in $models; do
   yaml="${model_dir}"/"${model}"/"${model}".yaml
+
+  # different naming scheme
+  if [[ "$model" == "Bertozzi_PNAS2020" ]]; then
+    yaml="${model_dir}"/"${model}"/problem.yaml
+  fi
+
   amici_model_dir=test_bmc/"${model}"
   mkdir -p "$amici_model_dir"
   cmd_import="amici_import_petab ${yaml} -o ${amici_model_dir} -n ${model} --flatten"
diff --git a/tests/benchmark-models/test_petab_benchmark.py b/tests/benchmark-models/test_petab_benchmark.py
index d0e2a6d46a..82d27f85eb 100644
--- a/tests/benchmark-models/test_petab_benchmark.py
+++ b/tests/benchmark-models/test_petab_benchmark.py
@@ -1,6 +1,5 @@
 """Tests for simulate_petab on PEtab benchmark problems."""
 
-import os
 from pathlib import Path
 
 import amici
@@ -9,39 +8,32 @@
 import petab.v1 as petab
 import pytest
 from amici.petab.petab_import import import_petab_problem
+import benchmark_models_petab
+
 
 # Absolute and relative tolerances for finite difference gradient checks.
 ATOL: float = 1e-3
 RTOL: float = 1e-2
 
 repo_root = Path(__file__).parent.parent.parent
-benchmark_path = repo_root / "Benchmark-Models-PEtab" / "Benchmark-Models"
-if not benchmark_path.exists():
-    benchmark_path = Path(os.environ["BENCHMARK_COLLECTION"])
 
 # reuse compiled models from test_benchmark_collection.sh
 benchmark_outdir = repo_root / "test_bmc"
-models = [
-    str(petab_path.stem)
-    for petab_path in benchmark_path.glob("*")
-    if petab_path.is_dir()
-    if str(petab_path.stem)
-    not in (
-        # excluded due to excessive runtime
-        "Bachmann_MSB2011",
-        "Chen_MSB2009",
-        "Froehlich_CellSystems2018",
-        "Raimundez_PCB2020",
-        "Lucarelli_CellSystems2018",
-        "Isensee_JCB2018",
-        "Beer_MolBioSystems2014",
-        "Alkan_SciSignal2018",
-        "Lang_PLOSComputBiol2024",
-        # excluded due to excessive numerical failures
-        "Crauste_CellSystems2017",
-        "Fujita_SciSignal2010",
-    )
-]
+models = set(benchmark_models_petab.MODELS) - {
+    # excluded due to excessive runtime
+    "Bachmann_MSB2011",
+    "Chen_MSB2009",
+    "Froehlich_CellSystems2018",
+    "Raimundez_PCB2020",
+    "Lucarelli_CellSystems2018",
+    "Isensee_JCB2018",
+    "Beer_MolBioSystems2014",
+    "Alkan_SciSignal2018",
+    "Lang_PLOSComputBiol2024",
+    # excluded due to excessive numerical failures
+    "Crauste_CellSystems2017",
+    "Fujita_SciSignal2010",
+}
 
 debug = False
 if debug:
@@ -75,9 +67,7 @@ def test_benchmark_gradient(model, scale):
         # only fail on linear scale
         pytest.skip()
 
-    petab_problem = petab.Problem.from_yaml(
-        benchmark_path / model / (model + ".yaml")
-    )
+    petab_problem = benchmark_models_petab.get_problem(model)
     petab.flatten_timepoint_specific_output_overrides(petab_problem)
 
     # Only compute gradient for estimated parameters.