diff --git a/benchmarks/gaussian.dx b/benchmarks/gaussian.dx
index 66e4e89cc..aa1eceb47 100644
--- a/benchmarks/gaussian.dx
+++ b/benchmarks/gaussian.dx
@@ -2,3 +2,7 @@ n = if dex_test_mode() then 100 else 1000 * 1000
 
 %bench "Gaussian"
 res = rand_vec n randn (new_key 0)
+>
+> Gaussian
+> Compile time: 40.102 ms
+> Run time:     8.517 us 	(based on 1 run)
diff --git a/benchmarks/jvp_matmul.dx b/benchmarks/jvp_matmul.dx
index adcff8074..acdb007ea 100644
--- a/benchmarks/jvp_matmul.dx
+++ b/benchmarks/jvp_matmul.dx
@@ -8,3 +8,7 @@ def mmp' {l m n} (m1:l=>m=>Float) (m2:m=>n=>Float) : l=>n=>Float =
 
 %bench "jvp_matmul"
 res = mmp' m1 m2
+>
+> jvp_matmul
+> Compile time: 82.255 ms
+> Run time:     7.298 us 	(based on 1 run)
diff --git a/benchmarks/matmul_big.dx b/benchmarks/matmul_big.dx
index 73ca88eed..383f9c570 100644
--- a/benchmarks/matmul_big.dx
+++ b/benchmarks/matmul_big.dx
@@ -5,3 +5,7 @@ m2 = rand_mat n n randn (new_key 1)
 
 %bench "matmul_big"
 res = m1 ** m2
+>
+> matmul_big
+> Compile time: 27.431 ms
+> Run time:     4.406 us 	(based on 1 run)
diff --git a/benchmarks/matmul_small.dx b/benchmarks/matmul_small.dx
index f4fe2a98f..1477ca86d 100644
--- a/benchmarks/matmul_small.dx
+++ b/benchmarks/matmul_small.dx
@@ -6,4 +6,8 @@ m2 = for i:(Fin width). rand_mat n n randn (new_key 1)
 
 %bench "matmul_small"
 res = for i. (m1.i ** m2.i)
+>
+> matmul_small
+> Compile time: 33.241 ms
+> Run time:     358.211 us 	(based on 1 run)
 
diff --git a/benchmarks/matvec_big.dx b/benchmarks/matvec_big.dx
index aecb6e858..bc8be0434 100644
--- a/benchmarks/matvec_big.dx
+++ b/benchmarks/matvec_big.dx
@@ -5,3 +5,7 @@ v = rand_vec n randn (new_key 1)
 
 %bench "matvec_big"
 res = m **. v
+>
+> matvec_big
+> Compile time: 23.898 ms
+> Run time:     4.682 us 	(based on 1 run)
diff --git a/benchmarks/matvec_small.dx b/benchmarks/matvec_small.dx
index 4556e98e2..b7a6d76e0 100644
--- a/benchmarks/matvec_small.dx
+++ b/benchmarks/matvec_small.dx
@@ -6,3 +6,7 @@ vs = for i:(Fin width). rand_vec n randn (new_key 1)
 
 %bench "matvec_small"
 res = for i. ms.i **. vs.i
+>
+> matvec_small
+> Compile time: 29.506 ms
+> Run time:     387.630 us 	(based on 1 run)
diff --git a/benchmarks/poly.dx b/benchmarks/poly.dx
index 856c5e91a..8dcd9419f 100644
--- a/benchmarks/poly.dx
+++ b/benchmarks/poly.dx
@@ -7,3 +7,7 @@ a = for i:(Fin n). n_to_f $ ordinal i
 
 %bench "poly"
 res = poly a
+>
+> poly
+> Compile time: 44.950 ms
+> Run time:     11.224 us 	(based on 1 run)
diff --git a/benchmarks/vjp_matmul.dx b/benchmarks/vjp_matmul.dx
index 6de63bb25..e0e687eab 100644
--- a/benchmarks/vjp_matmul.dx
+++ b/benchmarks/vjp_matmul.dx
@@ -8,3 +8,7 @@ def mmp' {l m n} (m1:l=>m=>Float) (m2:m=>n=>Float) : l=>n=>Float =
 
 %bench "jvp_matmul"
 res = mmp' m1 m2
+>
+> jvp_matmul
+> Compile time: 130.231 ms
+> Run time:     7.178 us 	(based on 1 run)
diff --git a/makefile b/makefile
index 07a57bdaf..1fa17b9be 100644
--- a/makefile
+++ b/makefile
@@ -216,7 +216,9 @@ test-names = uexpr-tests adt-tests type-tests eval-tests show-tests read-tests \
 
 doc-names = conditionals functions
 
-benchmark-names = fused_sum
+benchmark-names = \
+  fused_sum gaussian jvp_matmul matmul_big matmul_small matvec_big matvec_small \
+  poly vjp_matmul
 
 quine-test-targets = \
   $(test-names:%=run-tests/%) \