diff --git a/README.md b/README.md
index 013e700..04fffca 100644
--- a/README.md
+++ b/README.md
@@ -57,16 +57,19 @@ lama-cleaner --model=lama --device=cpu --port=8080
 
 Available arguments:
 
-| Name              | Description                                                                                              | Default  |
-| ----------------- | -------------------------------------------------------------------------------------------------------- | -------- |
-| --model           | lama/ldm/zits/mat/fcf/sd. See details in [Inpaint Model](#inpainting-model)                              | lama     |
-| --hf_access_token | stable-diffusion(sd) model need huggingface access token https://huggingface.co/docs/hub/security-tokens |          |
-| --device          | cuda or cpu                                                                                              | cuda     |
-| --port            | Port for backend flask web server                                                                        | 8080     |
-| --gui             | Launch lama-cleaner as a desktop application                                                             |          |
-| --gui_size        | Set the window size for the application                                                                  | 1200 900 |
-| --input           | Path to image you want to load by default                                                                | None     |
-| --debug           | Enable debug mode for flask web server                                                                   |          |
+| Name              | Description                                                                                                                   | Default  |
+|-------------------|-------------------------------------------------------------------------------------------------------------------------------| -------- |
+| --model           | lama/ldm/zits/mat/fcf/sd1.4 See details in [Inpaint Model](#inpainting-model)                                                 | lama     |
+| --hf_access_token | stable-diffusion(sd) model need [huggingface access token](https://huggingface.co/docs/hub/security-tokens) to download model |          |
+| --sd-run-local    | Once the model as downloaded, you can pass this arg and remove `--hf_access_token`                                            |          |
+| --sd-disable-nsfw | Disable stable-diffusion NSFW checker.                                                                                        |          |
+| --sd-cpu-textencoder | Always run stable-diffusion TextEncoder model on CPU.                                                                         |          |
+| --device          | cuda or cpu                                                                                                                   | cuda     |
+| --port            | Port for backend flask web server                                                                                             | 8080     |
+| --gui             | Launch lama-cleaner as a desktop application                                                                                  |          |
+| --gui_size        | Set the window size for the application                                                                                       | 1200 900 |
+| --input           | Path to image you want to load by default                                                                                     | None     |
+| --debug           | Enable debug mode for flask web server                                                                                        |          |
 
 ## Inpainting Model
 
diff --git a/lama_cleaner/model/sd.py b/lama_cleaner/model/sd.py
index 8aaeab8..a1cd881 100644
--- a/lama_cleaner/model/sd.py
+++ b/lama_cleaner/model/sd.py
@@ -70,7 +70,7 @@ class SD(InpaintModel):
     def init_model(self, device: torch.device, **kwargs):
         from .sd_pipeline import StableDiffusionInpaintPipeline
 
-        model_kwargs = {}
+        model_kwargs = {"local_files_only": kwargs['sd_run_local']}
         if kwargs['sd_disable_nsfw']:
             logger.info("Disable Stable Diffusion Model NSFW checker")
             model_kwargs.update(dict(
diff --git a/lama_cleaner/parse_args.py b/lama_cleaner/parse_args.py
index b876dde..e96bb64 100644
--- a/lama_cleaner/parse_args.py
+++ b/lama_cleaner/parse_args.py
@@ -20,13 +20,18 @@ def parse_args():
     parser.add_argument(
         "--sd-disable-nsfw",
         action="store_true",
-        help="Disable Stable Diffusion nsfw checker",
+        help="Disable Stable Diffusion NSFW checker",
     )
     parser.add_argument(
         "--sd-cpu-textencoder",
         action="store_true",
         help="Always run Stable Diffusion TextEncoder model on CPU",
     )
+    parser.add_argument(
+        "--sd-run-local",
+        action="store_true",
+        help="After first time Stable Diffusion model downloaded, you can add this arg and remove --hf_access_token",
+    )
     parser.add_argument("--device", default="cuda", type=str, choices=["cuda", "cpu"])
     parser.add_argument("--gui", action="store_true", help="Launch as desktop app")
     parser.add_argument(
@@ -48,7 +53,7 @@ def parse_args():
         if imghdr.what(args.input) is None:
             parser.error(f"invalid --input: {args.input} is not a valid image file")
 
-    if args.model.startswith("sd"):
+    if args.model.startswith("sd") and not args.sd_run_local:
         if not args.hf_access_token.startswith("hf_"):
             parser.error(
                 f"sd(stable-diffusion) model requires huggingface access token. Check how to get token from: https://huggingface.co/docs/hub/security-tokens"
diff --git a/lama_cleaner/server.py b/lama_cleaner/server.py
index 8fcd332..c72c782 100644
--- a/lama_cleaner/server.py
+++ b/lama_cleaner/server.py
@@ -220,6 +220,7 @@ def main(args):
         hf_access_token=args.hf_access_token,
         sd_disable_nsfw=args.sd_disable_nsfw,
         sd_cpu_textencoder=args.sd_cpu_textencoder,
+        sd_run_local=args.sd_run_local,
         callbacks=[diffuser_callback],
     )