diff --git a/README.md b/README.md
index 2b2496aa..25224f74 100644
--- a/README.md
+++ b/README.md
@@ -60,7 +60,7 @@ Unlike other distributed inference frameworks, exo does not use a master-worker
 
 Exo supports different [partitioning strategies](exo/topology/partitioning_strategy.py) to split up a model across devices. The default partitioning strategy is [ring memory weighted partitioning](exo/topology/ring_memory_weighted_partitioning_strategy.py). This runs an inference in a ring where each device runs a number of model layers proportional to the memory of the device.
 
-!["A screenshot of exo running 5 nodes](docs/exo-screenshot.png)
+!["A screenshot of exo running 5 nodes](docs/exo-screenshot.jpg)
 
 ## Installation
 
diff --git a/docs/exo-screenshot.jpg b/docs/exo-screenshot.jpg
new file mode 100644
index 00000000..ace09a46
Binary files /dev/null and b/docs/exo-screenshot.jpg differ
diff --git a/docs/exo-screenshot.png b/docs/exo-screenshot.png
deleted file mode 100644
index 19963ad2..00000000
--- a/docs/exo-screenshot.png
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:561ec71a226a154503b1d70553c9d57c7cd45dbb3bb0e1244ed5b00edbf0523d
-size 479724
diff --git a/docs/ring-topology.png b/docs/ring-topology.png
deleted file mode 100644
index 6c7d8fd9..00000000
--- a/docs/ring-topology.png
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3f57b11f2d3aefb3887c5266994c4b4335501830c77a6a53fa6901c8725d0f6c
-size 55857