diff --git a/scripts/deforum_helpers/animation_key_frames.py b/scripts/deforum_helpers/animation_key_frames.py
index ca4aa4310..af3126252 100644
--- a/scripts/deforum_helpers/animation_key_frames.py
+++ b/scripts/deforum_helpers/animation_key_frames.py
@@ -87,6 +87,21 @@ def __init__(self, anim_args, controlnet_args):
                 self.schedules[output_key] = self.fi.parse_inbetweens(getattr(controlnet_args, input_key), input_key)
                 setattr(self, output_key, self.schedules[output_key])
 
+class AnimateDiffKeys():
+    def __init__(self, animatediff_args, anim_args):
+        self.fi = FrameInterpolater(anim_args.max_frames)
+        self.enable = animatediff_args.animatediff_enabled
+        self.model = animatediff_args.animatediff_model
+        self.activation_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_activation_schedule, 'activation_schedule')
+        self.motion_lora_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_motion_lora_schedule, 'motion_lora_schedule', is_single_string = True)
+        self.video_length_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_video_length_schedule, 'video_length_schedule')
+        self.batch_size_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_batch_size_schedule, 'batch_size_schedule')
+        self.stride_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_stride_schedule, 'stride_schedule')
+        self.overlap_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_overlap_schedule, 'overlap_schedule')
+        self.latent_scale_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_latent_scale_schedule, 'latent_scale_schedule')
+        self.latent_power_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_latent_power_schedule, 'latent_power_schedule')
+        self.closed_loop_schedule_series = self.fi.parse_inbetweens(animatediff_args.animatediff_closed_loop_schedule, 'closed_loop_schedule', is_single_string = True)
+
 class LooperAnimKeys():
     def __init__(self, loop_args, anim_args, seed):
         self.fi = FrameInterpolater(anim_args.max_frames, seed)
diff --git a/scripts/deforum_helpers/args.py b/scripts/deforum_helpers/args.py
index 98257597f..d354c4372 100644
--- a/scripts/deforum_helpers/args.py
+++ b/scripts/deforum_helpers/args.py
@@ -23,6 +23,7 @@
 import modules.shared as sh
 from modules.processing import get_fixed_seed
 from .defaults import get_guided_imgs_default_json, mask_fill_choices, get_samplers_list
+from .deforum_animatediff import animatediff_component_names
 from .deforum_controlnet import controlnet_component_names
 from .general_utils import get_os, substitute_placeholders
 
@@ -1119,7 +1120,7 @@ def DeforumOutputArgs():
 
 def get_component_names():
     return ['override_settings_with_file', 'custom_settings_file', *DeforumAnimArgs().keys(), 'animation_prompts', 'animation_prompts_positive', 'animation_prompts_negative',
-            *DeforumArgs().keys(), *DeforumOutputArgs().keys(), *ParseqArgs().keys(), *LoopArgs().keys(), *controlnet_component_names()]
+            *DeforumArgs().keys(), *DeforumOutputArgs().keys(), *ParseqArgs().keys(), *LoopArgs().keys(), *animatediff_component_names(), *controlnet_component_names()]
 
 def get_settings_component_names():
     return [name for name in get_component_names()]
@@ -1139,13 +1140,14 @@ def process_args(args_dict_main, run_id):
     video_args = SimpleNamespace(**{name: args_dict_main[name] for name in DeforumOutputArgs()})
     parseq_args = SimpleNamespace(**{name: args_dict_main[name] for name in ParseqArgs()})
     loop_args = SimpleNamespace(**{name: args_dict_main[name] for name in LoopArgs()})
+    animatediff_args = SimpleNamespace(**{name: args_dict_main[name] for name in animatediff_component_names()})
     controlnet_args = SimpleNamespace(**{name: args_dict_main[name] for name in controlnet_component_names()})
 
     root.animation_prompts = json.loads(args_dict_main['animation_prompts'])
 
     args_loaded_ok = True
     if override_settings_with_file:
-        args_loaded_ok = load_args(args_dict_main, args, anim_args, parseq_args, loop_args, controlnet_args, video_args, custom_settings_file, root, run_id)
+        args_loaded_ok = load_args(args_dict_main, args, anim_args, parseq_args, loop_args, animatediff_args, controlnet_args, video_args, custom_settings_file, root, run_id)
 
     positive_prompts = args_dict_main['animation_prompts_positive']
     negative_prompts = args_dict_main['animation_prompts_negative']
@@ -1184,4 +1186,4 @@ def process_args(args_dict_main, run_id):
     default_img = default_img.resize((args.W,args.H))
     root.default_img = default_img
 
-    return args_loaded_ok, root, args, anim_args, video_args, parseq_args, loop_args, controlnet_args
+    return args_loaded_ok, root, args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args
diff --git a/scripts/deforum_helpers/deforum_animatediff.py b/scripts/deforum_helpers/deforum_animatediff.py
new file mode 100644
index 000000000..5091586d1
--- /dev/null
+++ b/scripts/deforum_helpers/deforum_animatediff.py
@@ -0,0 +1,262 @@
+# Copyright (C) 2023 Deforum LLC
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as published by
+# the Free Software Foundation, version 3 of the License.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with this program. If not, see <https://www.gnu.org/licenses/>.
+
+# Contact the authors: https://deforum.github.io/
+
+# This helper script is responsible for AnimateDiff/Deforum integration
+# https://github.com/continue-revolution/sd-webui-animatediff — animatediff repo
+
+import os
+import copy
+import gradio as gr
+import scripts
+from PIL import Image
+import numpy as np
+import importlib
+import shutil
+from modules import scripts, shared
+from .deforum_controlnet_gradio import hide_ui_by_cn_status, hide_file_textboxes, ToolButton
+from .general_utils import count_files_in_folder, clean_gradio_path_strings  # TODO: do it another way
+from .video_audio_utilities import vid2frames, convert_image
+from .animation_key_frames import AnimateDiffKeys
+from .load_images import load_image
+from .general_utils import debug_print
+from modules.shared import opts, cmd_opts, state, sd_model
+
+import modules.paths as ph
+
+#self.last_frame = last_frame
+#self.latent_power_last = latent_power_last
+#self.latent_scale_last = latent_scale_last
+
+cnet = None
+
+def find_animatediff():
+    global cnet
+    if cnet: return cnet
+    try:
+        cnet = importlib.import_module('extensions.sd-webui-animatediff.scripts', 'animatediff')
+    except:
+        try:
+            cnet = importlib.import_module('extensions-builtin.sd-webui-animatediff.scripts', 'animatediff')
+        except:
+            pass
+    if cnet:
+        print(f"\033[0;32m*Deforum AnimateDiff support: enabled*\033[0m")
+        return True
+    return None
+
+def is_animatediff_enabled(animatediff_args):
+    if getattr(animatediff_args, f'animatediff_enabled', False):
+        return True
+    return False
+
+def animatediff_infotext():
+    return """**Experimental!**
+Requires the <a style='color:SteelBlue;' target='_blank' href='https://github.com/continue-revolution/sd-webui-animatediff'>AnimateDiff</a> extension to be installed.</p>
+"""
+
+def animatediff_component_names_raw():
+    return [
+        'enabled', 'model', 'activation_schedule',
+        'motion_lora_schedule',
+        'video_length_schedule',
+        'batch_size_schedule',
+        'stride_schedule',
+        'overlap_schedule',
+        'latent_power_schedule', 'latent_scale_schedule',
+        'closed_loop_schedule'
+    ]
+
+def animatediff_component_names():
+    if not find_animatediff():
+        return []
+
+    return [f'animatediff_{i}' for i in animatediff_component_names_raw()]
+
+def setup_animatediff_ui_raw():
+
+    cnet = find_animatediff()
+
+    model_dir = shared.opts.data.get("animatediff_model_path", os.path.join(scripts.basedir(), "model"))
+
+    if not os.path.isdir(model_dir):
+        os.mkdir(model_dir)
+
+    cn_models = [f for f in os.listdir(model_dir) if f != ".gitkeep"]
+
+    def refresh_all_models(*inputs):
+        new_model_list = [
+            f for f in os.listdir(model_dir) if f != ".gitkeep"
+        ]
+        dd = inputs[0]
+        if dd in new_model_list:
+            selected = dd
+        elif len(new_model_list) > 0:
+            selected = new_model_list[0]
+        else:
+            selected = None
+        return gr.Dropdown.update(choices=new_model_list, value=selected)
+
+    refresh_symbol = '\U0001f504'  # 🔄
+    switch_values_symbol = '\U000021C5'  # ⇅
+    infotext_fields = []
+
+    # TODO: unwrap
+    def create_model_in_tab_ui(cn_id):
+        with gr.Row():
+            gr.Markdown('Note: AnimateDiff will work only if you have ControlNet installed as well')
+            enabled = gr.Checkbox(label="Enable AnimateDiff", value=False, interactive=True)
+        with gr.Row(visible=False) as mod_row:
+            model = gr.Dropdown(cn_models, label=f"Motion module", value="None", interactive=True, tooltip="Choose which motion module will be injected into the generation process.")
+            refresh_models = ToolButton(value=refresh_symbol)
+            refresh_models.click(refresh_all_models, model, model)
+        with gr.Row(visible=False) as inforow:
+            gr.Markdown('**Important!** This schedule sets up when AnimateDiff should run on the generated N previous frames. At the moment this is made with binary values: when the expression value is 0, it will make a pass, otherwise normal Deforum frames will be made')
+        with gr.Row(visible=False) as activation_row:
+            activation_schedule = gr.Textbox(label="AnimateDiff activation schedule", lines=1, value='0:(1), 2:((t-1) % 16)', interactive=True)
+        gr.Markdown('Internal AnimateDiff settings, see its script in normal tabs')
+        with gr.Row(visible=False) as motion_lora_row:
+            motion_lora_schedule = gr.Textbox(label="Motion lora schedule (not supported atm, stay tuned!)", lines=1, value='0:("")', interactive=True)
+        with gr.Row(visible=False) as length_row:
+            video_length_schedule = gr.Textbox(label="N-back video length schedule", lines=1, value='0:(16)', interactive=True)
+        with gr.Row(visible=False) as window_row:
+            batch_size_schedule = gr.Textbox(label="Batch size", lines=1, value='0:(16)', interactive=True)
+        with gr.Row(visible=False) as stride_row:
+            stride_schedule = gr.Textbox(label="Stride", lines=1, value='0:(1)', interactive=True)
+        with gr.Row(visible=False) as overlap_row:
+            overlap_schedule = gr.Textbox(label="Overlap", lines=1, value='0:(-1)', interactive=True)
+        with gr.Row(visible=False) as latent_power_row:
+            latent_power_schedule = gr.Textbox(label="Latent power schedule", lines=1, value='0:(1)', interactive=True)
+        with gr.Row(visible=False) as latent_scale_row:
+            latent_scale_schedule = gr.Textbox(label="Latent scale schedule", lines=1, value='0:(32)', interactive=True)
+        with gr.Row(visible=False) as rp_row:
+            closed_loop_schedule = gr.Textbox(label="Closed loop", lines=1, value='0:("R-P")', interactive=True)
+        hide_output_list = [enabled, inforow, activation_row, motion_lora_row, mod_row, length_row, window_row, stride_row, overlap_row, latent_power_row, latent_scale_row, rp_row]
+        for cn_output in hide_output_list:
+            enabled.change(fn=hide_ui_by_cn_status, inputs=enabled, outputs=cn_output)
+
+        infotext_fields.extend([
+            (model, f"AnimateDiff Model"),
+        ])
+
+        return {key: value for key, value in locals().items() if key in 
+            animatediff_component_names_raw()
+        }
+
+    with gr.TabItem('AnimateDiff'):
+        gr.HTML(animatediff_infotext())
+        model_params = create_model_in_tab_ui(0)
+
+        for key, value in model_params.items():
+            locals()[f"animatediff_{key}"] = value
+
+    return locals()
+
+def setup_animatediff_ui():
+    if not find_animatediff():
+        gr.HTML("""<a style='target='_blank' href='https://github.com/continue-revolution/sd-webui-animatediff'>AnimateDiff not found. Please install it :)</a>""", elem_id='animatediff_not_found_html_msg')
+        return {}
+
+    try:
+        return setup_animatediff_ui_raw()
+    except Exception as e:
+        print(f"'AnimateDiff UI setup failed with error: '{e}'!")
+        gr.HTML(f"""
+                Failed to setup AnimateDiff UI, check the reason in your commandline log. Please, downgrade your AnimateDiff extension to <a style='color:Orange;' target='_blank' href='https://github.com/continue-revolution/sd-webui-animatediff/archive/b192a2551a5ed66d4a3ce58d5d19a8872abc87ca.zip'>b192a2551a5ed66d4a3ce58d5d19a8872abc87ca</a> and report the problem <a style='color:Orange;' target='_blank' href='https://github.com/deforum-art/sd-webui-deforum'>here</a> (Deforum) or <a style='color:Orange;' target='_blank' href='https://github.com/continue-revolution/sd-webui-animatediff'>here</a> (AnimateDiff).
+                """, elem_id='animatediff_not_found_html_msg')
+        return {}
+
+def find_animatediff_script(prev_always_on_scripts):
+    animatediff_script = next((script for script in prev_always_on_scripts if "animatediff" in script.title().lower()), None)
+    if not animatediff_script:
+        raise Exception("AnimateDiff script not found.")
+    return animatediff_script
+
+def get_animatediff_temp_dir(args):
+    return os.path.join(args.outdir, 'animatediff_temp')
+
+def need_animatediff(animatediff_args):
+    return find_animatediff() is not None and is_animatediff_enabled(animatediff_args)
+
+def seed_animatediff(p, prev_always_on_scripts, animatediff_args, args, anim_args, root, frame_idx):
+    if not need_animatediff(animatediff_args):
+        return
+
+    keys = AnimateDiffKeys(animatediff_args, anim_args) # if not parseq_adapter.use_parseq else parseq_adapter.cn_keys
+    
+    # Will do the back-render only on target frames
+    if int(keys.activation_schedule_series[frame_idx]) != 0:
+        return
+    
+    video_length = int(keys.video_length_schedule_series[frame_idx])
+    assert video_length > 1
+
+    # Managing the frames to be fed into AD:
+    # Create a temporal directory
+    animatediff_temp_dir = get_animatediff_temp_dir(args)
+    if os.path.exists(animatediff_temp_dir):
+        shutil.rmtree(animatediff_temp_dir)
+    os.makedirs(animatediff_temp_dir)
+    # Copy the frames (except for the one which is being CN-made) into that dir
+    for offset in range(video_length - 1):
+        filename = f"{root.timestring}_{frame_idx - offset - 1:09}.png"
+        Image.open(os.path.join(args.outdir, filename)).save(os.path.join(animatediff_temp_dir, f"{offset:09}.png"), "PNG")
+
+    animatediff_script = find_animatediff_script(prev_always_on_scripts)
+    # let's put it before ControlNet to cause less problems
+    p.is_api = True # to parse the params internally
+    p.scripts.alwayson_scripts = [animatediff_script] + p.scripts.alwayson_scripts
+
+    args_dict = {
+        'model': keys.model,   # Motion module
+        'format': ['PNG', 'Frame'],      # Save format, 'GIF' | 'MP4' | 'PNG' | 'WEBP' | 'WEBM' | 'TXT' | 'Frame'
+        'enable': keys.enable,         # Enable AnimateDiff
+        'video_length': video_length,     # Number of frames
+        'fps': 8,               # FPS - don't care
+        'loop_number': 0,       # Display loop number
+        'closed_loop': keys.closed_loop_schedule_series[frame_idx],   # Closed loop, 'N' | 'R-P' | 'R+P' | 'A'
+        'batch_size': int(keys.batch_size_schedule_series[frame_idx]),       # Context batch size
+        'stride': int(keys.stride_schedule_series[frame_idx]),            # Stride 
+        'overlap': int(keys.overlap_schedule_series[frame_idx]),          # Overlap
+        'interp': 'Off',        # Frame interpolation, 'Off' | 'FILM' - don't care
+        'interp_x': 10,          # Interp X - don't care
+        'video_source': '',  # We don't use a video
+        'video_path': animatediff_temp_dir, # Path with our selected video_length input frames
+        'latent_power': keys.latent_power_schedule_series[frame_idx],      # Latent power
+        'latent_scale': keys.latent_scale_schedule_series[frame_idx],     # Latent scale
+        'last_frame': None,     # Optional last frame
+        'latent_power_last': 1, # Optional latent power for last frame
+        'latent_scale_last': 32,# Optional latent scale for last frame
+        'request_id': ''        # Optional request id. If provided, outputs will have request id as filename suffix
+    }
+
+    args = [None] * 10 + [args_dict] # HACK hardcoded args offset
+
+    p.script_args_value = args + p.script_args_value
+
+def reap_animatediff(images, animatediff_args, args, root, frame_idx):
+    if not need_animatediff(animatediff_args):
+        return
+    
+    animatediff_temp_dir = get_animatediff_temp_dir(args)
+    assert os.path.exists(animatediff_temp_dir)
+
+    for offset in range(len(images)):
+        frame = images[-offset-1]
+        cur_frame_idx = frame_idx - offset
+
+        # overwrite the results
+        filename = f"{root.timestring}_{cur_frame_idx:09}.png"
+        frame.save(os.path.join(args.outdir, filename), "PNG")
diff --git a/scripts/deforum_helpers/deforum_controlnet.py b/scripts/deforum_helpers/deforum_controlnet.py
index cef20cbc8..729b85036 100644
--- a/scripts/deforum_helpers/deforum_controlnet.py
+++ b/scripts/deforum_helpers/deforum_controlnet.py
@@ -38,6 +38,9 @@
 max_models = shared.opts.data.get("control_net_unit_count", shared.opts.data.get("control_net_max_models_num", 5))
 num_of_models = 5 if max_models <= 5 else max_models
 
+# AnimateDiff support (it requires ControlNet anyway)
+from .deforum_animatediff import seed_animatediff, is_animatediff_enabled
+
 def find_controlnet():
     global cnet
     if cnet: return cnet
@@ -217,7 +220,8 @@ def controlnet_component_names():
         'processor_res', 'threshold_a', 'threshold_b', 'resize_mode', 'control_mode', 'loopback_mode'
     ]]
 
-def process_with_controlnet(p, args, anim_args, controlnet_args, root, parseq_adapter, is_img2img=True, frame_idx=0):
+def process_with_controlnet(p, args, anim_args, controlnet_args, animatediff_args, root, parseq_adapter, is_img2img=True, frame_idx=0):
+    p.do_not_save_grid = True
     CnSchKeys = ControlNetKeys(anim_args, controlnet_args) if not parseq_adapter.use_parseq else parseq_adapter.cn_keys
 
     def read_cn_data(cn_idx):
@@ -264,7 +268,7 @@ def read_cn_data(cn_idx):
 
     cn_inputframes_list = [os.path.join(args.outdir, f'controlnet_{i}_inputframes') for i in range(1, num_of_models + 1)]
 
-    if not any(os.path.exists(cn_inputframes) for cn_inputframes in cn_inputframes_list) and not any_loopback_mode:
+    if not any(os.path.exists(cn_inputframes) for cn_inputframes in cn_inputframes_list) and not any_loopback_mode and not is_animatediff_enabled(animatediff_args):
         print(f'\033[33mNeither the base nor the masking frames for ControlNet were found. Using the regular pipeline\033[0m')
 
     # Remove all scripts except controlnet.
@@ -284,11 +288,15 @@ def read_cn_data(cn_idx):
     #
     p.scripts = copy.copy(scripts.scripts_img2img if is_img2img else scripts.scripts_txt2img)
     controlnet_script = find_controlnet_script(p)
+    prev_always_on_scripts = p.scripts.alwayson_scripts
     p.scripts.alwayson_scripts =  [controlnet_script]
     # Filling the list with None is safe because only the length will be considered,
     # and all cn args will be replaced.
     p.script_args_value = [None] * controlnet_script.args_to
 
+    # Basically, launch AD on a number of previous frames once it hits the seed time
+    seed_animatediff(p, prev_always_on_scripts, animatediff_args, args, anim_args, root, frame_idx)
+
     def create_cnu_dict(cn_args, prefix, img_np, mask_np, frame_idx, CnSchKeys):
 
         keys = [
diff --git a/scripts/deforum_helpers/generate.py b/scripts/deforum_helpers/generate.py
index fe48055e3..261a63987 100644
--- a/scripts/deforum_helpers/generate.py
+++ b/scripts/deforum_helpers/generate.py
@@ -36,6 +36,7 @@
 from types import SimpleNamespace
 
 from .general_utils import debug_print
+from .deforum_animatediff import reap_animatediff, is_animatediff_enabled
 
 def load_mask_latent(mask_input, shape):
     # mask_input (str or PIL Image.Image): Path to the mask image or a PIL Image object
@@ -70,14 +71,14 @@ def pairwise_repl(iterable):
     next(b, None)
     return zip(a, b)
 
-def generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter,  frame=0, sampler_name=None):
+def generate(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter,  frame=0, sampler_name=None):
     if state.interrupted:
         return None
 
     if args.reroll_blank_frames == 'ignore':
-        return generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame, sampler_name)
+        return generate_inner(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame, sampler_name)
 
-    image, caught_vae_exception = generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame, sampler_name)
+    image, caught_vae_exception = generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame, sampler_name)
 
     if caught_vae_exception or not image.getbbox():
         patience = args.reroll_patience
@@ -86,7 +87,7 @@ def generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_ada
             while caught_vae_exception or not image.getbbox():
                 print("Rerolling with +1 seed...")
                 args.seed += 1
-                image, caught_vae_exception = generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame, sampler_name)
+                image, caught_vae_exception = generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame, sampler_name)
                 patience -= 1
                 if patience == 0:
                     print("Rerolling with +1 seed failed for 10 iterations! Try setting webui's precision to 'full' and if it fails, please report this to the devs! Interrupting...")
@@ -100,12 +101,12 @@ def generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_ada
             return None
     return image
 
-def generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame=0, sampler_name=None):
+def generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame=0, sampler_name=None):
     if cmd_opts.disable_nan_check:
-        image = generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame, sampler_name)
+        image = generate_inner(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame, sampler_name)
     else:
         try:
-            image = generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame, sampler_name)
+            image = generate_inner(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame, sampler_name)
         except Exception as e:
             if "A tensor with all NaNs was produced in VAE." in repr(e):
                 print(e)
@@ -114,7 +115,7 @@ def generate_with_nans_check(args, keys, anim_args, loop_args, controlnet_args,
                 raise e
     return image, False
 
-def generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame=0, sampler_name=None):
+def generate_inner(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame=0, sampler_name=None):
     # Setup the pipeline
     p = get_webui_sd_pipeline(args, root)
     p.prompt, p.negative_prompt = split_weighted_subprompts(args.prompt, frame, anim_args.max_frames)
@@ -234,8 +235,8 @@ def generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, pars
 
             print_combined_table(args, anim_args, p_txt, keys, frame)  # print dynamic table to cli
 
-            if is_controlnet_enabled(controlnet_args):
-                process_with_controlnet(p_txt, args, anim_args, controlnet_args, root, parseq_adapter, is_img2img=False, frame_idx=frame)
+            if is_controlnet_enabled(controlnet_args) or is_animatediff_enabled(animatediff_args):
+                process_with_controlnet(p_txt, args, anim_args, controlnet_args, animatediff_args, root, parseq_adapter, is_img2img=False, frame_idx=frame)
 
             with A1111OptionsOverrider({"control_net_detectedmap_dir" : os.path.join(args.outdir, "controlnet_detected_map")}):
                 processed = processing.process_images(p_txt)
@@ -276,8 +277,8 @@ def generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, pars
         if args.motion_preview_mode:
             processed = mock_process_images(args, p, init_image)
         else:
-            if is_controlnet_enabled(controlnet_args):
-                process_with_controlnet(p, args, anim_args, controlnet_args, root, parseq_adapter, is_img2img=True, frame_idx=frame)
+            if is_controlnet_enabled(controlnet_args) or is_animatediff_enabled(animatediff_args):
+                process_with_controlnet(p, args, anim_args, controlnet_args, animatediff_args, root, parseq_adapter, is_img2img=True, frame_idx=frame)
             
             with A1111OptionsOverrider({"control_net_detectedmap_dir" : os.path.join(args.outdir, "controlnet_detected_map")}):
                 processed = processing.process_images(p)
@@ -287,9 +288,12 @@ def generate_inner(args, keys, anim_args, loop_args, controlnet_args, root, pars
         root.initial_info = processed.info
 
     if root.first_frame is None:
-        root.first_frame = processed.images[0]
+        root.first_frame = processed.images[-1]
 
-    results = processed.images[0]
+    results = processed.images[-1] # AD uses ascending order, so we need to get the last frame
+
+    if len(processed.images) > 1:
+        reap_animatediff(processed.images, animatediff_args, args, root, frame)
 
     return results
 
diff --git a/scripts/deforum_helpers/render.py b/scripts/deforum_helpers/render.py
index 964498c3a..247a5ef80 100644
--- a/scripts/deforum_helpers/render.py
+++ b/scripts/deforum_helpers/render.py
@@ -52,11 +52,15 @@
 
 from deforum_api import JobStatusTracker
 
-def render_animation(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root):
+def render_animation(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root):
 
     # initialise Parseq adapter
+    # TODO: @rewbs
     parseq_adapter = ParseqAdapter(parseq_args, anim_args, video_args, controlnet_args, loop_args)
 
+    if animatediff_args.animatediff_enabled:
+        print("*Rendering with AnimateDiff turned on. (Experimental!)*")        
+
     if opts.data.get("deforum_save_gen_info_as_srt", False):  # create .srt file and set timeframe mechanism using FPS
         srt_filename = os.path.join(args.outdir, f"{root.timestring}.srt")
         srt_frame_duration = init_srt_file(srt_filename, video_args.fps)
@@ -93,7 +97,7 @@ def render_animation(args, anim_args, video_args, parseq_args, loop_args, contro
     print(f"Saving animation frames to:\n{args.outdir}")
 
     # save settings.txt file for the current run
-    save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, video_args, root)
+    save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, animatediff_args, video_args, root)
 
     # resume from timestring
     if anim_args.resume_from_timestring:
@@ -547,7 +551,7 @@ def render_animation(args, anim_args, video_args, parseq_args, loop_args, contro
             args.seed = random.randint(0, 2 ** 32 - 1)
             print(f"Optical flow redo is diffusing and warping using {optical_flow_redo_generation} and seed {args.seed} optical flow before generation.")
 
-            disposable_image = generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
+            disposable_image = generate(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
             disposable_image = cv2.cvtColor(np.array(disposable_image), cv2.COLOR_RGB2BGR)
             disposable_flow = get_flow_from_images(prev_img, disposable_image, optical_flow_redo_generation, raft_model)
             disposable_image = cv2.cvtColor(disposable_image, cv2.COLOR_BGR2RGB)
@@ -563,7 +567,7 @@ def render_animation(args, anim_args, video_args, parseq_args, loop_args, contro
             for n in range(0, int(anim_args.diffusion_redo)):
                 print(f"Redo generation {n + 1} of {int(anim_args.diffusion_redo)} before final generation")
                 args.seed = random.randint(0, 2 ** 32 - 1)
-                disposable_image = generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
+                disposable_image = generate(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
                 disposable_image = cv2.cvtColor(np.array(disposable_image), cv2.COLOR_RGB2BGR)
                 # color match on last one only
                 if n == int(anim_args.diffusion_redo):
@@ -574,7 +578,7 @@ def render_animation(args, anim_args, video_args, parseq_args, loop_args, contro
             gc.collect()
 
         # generation
-        image = generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
+        image = generate(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
 
         if image is None:
             break
diff --git a/scripts/deforum_helpers/render_modes.py b/scripts/deforum_helpers/render_modes.py
index 46730c991..388ecfa76 100644
--- a/scripts/deforum_helpers/render_modes.py
+++ b/scripts/deforum_helpers/render_modes.py
@@ -30,7 +30,7 @@
 from .save_images import save_image
 from .settings import save_settings_from_animation_run
 
-def render_input_video(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root):
+def render_input_video(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root):
     # create a folder for the video input frames to live in
     video_in_frame_path = os.path.join(args.outdir, 'inputframes') 
     os.makedirs(video_in_frame_path, exist_ok=True)
@@ -61,10 +61,10 @@ def render_input_video(args, anim_args, video_args, parseq_args, loop_args, cont
         args.use_mask = True
         args.overlay_mask = True
 
-    render_animation(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root)
+    render_animation(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root)
 
 # Modified a copy of the above to allow using masking video with out a init video.
-def render_animation_with_video_mask(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root):
+def render_animation_with_video_mask(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root):
     # create a folder for the video input frames to live in
     mask_in_frame_path = os.path.join(args.outdir, 'maskframes') 
     os.makedirs(mask_in_frame_path, exist_ok=True)
@@ -80,7 +80,7 @@ def render_animation_with_video_mask(args, anim_args, video_args, parseq_args, l
     #args.use_init = True
     print(f"Loading {anim_args.max_frames} input frames from {mask_in_frame_path} and saving video frames to {args.outdir}")
 
-    render_animation(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root)
+    render_animation(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root)
 
 def get_parsed_value(value, frame_idx, max_f):
     pattern = r'`.*?`'
@@ -93,7 +93,7 @@ def get_parsed_value(value, frame_idx, max_f):
         parsed_value = parsed_value.replace(matched_string, str(value))
     return parsed_value
 
-def render_interpolation(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root):
+def render_interpolation(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root):
 
     # use parseq if manifest is provided
     parseq_adapter = ParseqAdapter(parseq_args, anim_args, video_args, controlnet_args, loop_args)
@@ -106,7 +106,7 @@ def render_interpolation(args, anim_args, video_args, parseq_args, loop_args, co
     print(f"Saving interpolation animation frames to {args.outdir}")
 
     # save settings.txt file for the current run
-    save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, video_args, root)
+    save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, animatediff_args, video_args, root)
         
     # Compute interpolated prompts
     if parseq_adapter.manages_prompts():
@@ -162,7 +162,7 @@ def render_interpolation(args, anim_args, video_args, parseq_args, loop_args, co
         args.seed = int(keys.seed_schedule_series[frame_idx]) if (args.seed_behavior == 'schedule' or parseq_adapter.manages_seed()) else args.seed
         opts.data["CLIP_stop_at_last_layers"] = scheduled_clipskip if scheduled_clipskip is not None else opts.data["CLIP_stop_at_last_layers"]
 
-        image = generate(args, keys, anim_args, loop_args, controlnet_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
+        image = generate(args, keys, anim_args, loop_args, controlnet_args, animatediff_args, root, parseq_adapter, frame_idx, sampler_name=scheduled_sampler_name)
         filename = f"{root.timestring}_{frame_idx:09}.png"
 
         save_image(image, 'PIL', filename, args, video_args, root)
diff --git a/scripts/deforum_helpers/run_deforum.py b/scripts/deforum_helpers/run_deforum.py
index ca25d152b..5065c81eb 100644
--- a/scripts/deforum_helpers/run_deforum.py
+++ b/scripts/deforum_helpers/run_deforum.py
@@ -71,7 +71,7 @@ def run_deforum(*args):
         args_dict['self'] = None
         args_dict['p'] = p
         try:
-            args_loaded_ok, root, args, anim_args, video_args, parseq_args, loop_args, controlnet_args = process_args(args_dict, i)
+            args_loaded_ok, root, args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args = process_args(args_dict, i)
         except Exception as e:
             JobStatusTracker().fail_job(job_id, error_type="TERMINAL", message="Invalid arguments.")
             print("\n*START OF TRACEBACK*")
@@ -111,13 +111,13 @@ def run_deforum(*args):
             JobStatusTracker().update_output_info(job_id, outdir=args.outdir, timestring=root.timestring)
             if anim_args.animation_mode == '2D' or anim_args.animation_mode == '3D':
                 if anim_args.use_mask_video: 
-                    render_animation_with_video_mask(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root)  # allow mask video without an input video
+                    render_animation_with_video_mask(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root)  # allow mask video without an input video
                 else:    
-                    render_animation(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root)
+                    render_animation(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root)
             elif anim_args.animation_mode == 'Video Input':
-                render_input_video(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root)#TODO: prettify code
+                render_input_video(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root)#TODO: prettify code
             elif anim_args.animation_mode == 'Interpolation':
-                render_interpolation(args, anim_args, video_args, parseq_args, loop_args, controlnet_args, root)
+                render_interpolation(args, anim_args, video_args, parseq_args, loop_args, animatediff_args, controlnet_args, root)
             else:
                 print('Other modes are not available yet!')
         except Exception as e:
@@ -221,7 +221,7 @@ def run_deforum(*args):
 
         if shared.opts.data.get("deforum_enable_persistent_settings", False):
             persistent_sett_path = shared.opts.data.get("deforum_persistent_settings_path")
-            save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, video_args, root, persistent_sett_path)
+            save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, animatediff_args, controlnet_args, video_args, root, persistent_sett_path)
 
         # Close the pipeline, not to interfere with ControlNet
         try:
diff --git a/scripts/deforum_helpers/settings.py b/scripts/deforum_helpers/settings.py
index bd1132d9b..214f5821f 100644
--- a/scripts/deforum_helpers/settings.py
+++ b/scripts/deforum_helpers/settings.py
@@ -58,7 +58,7 @@ def load_args(args_dict_main, args, anim_args, parseq_args, loop_args, controlne
         return True
 
 # save settings function that get calls when run_deforum is being called
-def save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, video_args, root, full_out_file_path = None):
+def save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, controlnet_args, animatediff_args, video_args, root, full_out_file_path = None):
     if full_out_file_path:
         args.__dict__["seed"] = root.raw_seed
         args.__dict__["batch_name"] = root.raw_batch_name
@@ -69,7 +69,7 @@ def save_settings_from_animation_run(args, anim_args, parseq_args, loop_args, co
     settings_filename = full_out_file_path if full_out_file_path else os.path.join(args.outdir, f"{root.timestring}_settings.txt")
     with open(settings_filename, "w+", encoding="utf-8") as f:
         s = {}
-        for d in (args.__dict__, anim_args.__dict__, parseq_args.__dict__, loop_args.__dict__, controlnet_args.__dict__, video_args.__dict__):
+        for d in (args.__dict__, anim_args.__dict__, parseq_args.__dict__, loop_args.__dict__, controlnet_args.__dict__, animatediff_args.__dict__, video_args.__dict__):
             s.update({k: v for k, v in d.items() if k not in exclude_keys})
         s["sd_model_name"] = sh.sd_model.sd_checkpoint_info.name
         s["sd_model_hash"] = sh.sd_model.sd_checkpoint_info.hash
diff --git a/scripts/deforum_helpers/ui_left.py b/scripts/deforum_helpers/ui_left.py
index aa1886372..57b4df03a 100644
--- a/scripts/deforum_helpers/ui_left.py
+++ b/scripts/deforum_helpers/ui_left.py
@@ -20,6 +20,7 @@
 from .gradio_funcs import change_css, handle_change_functions
 from .args import DeforumArgs, DeforumAnimArgs, ParseqArgs, DeforumOutputArgs, RootArgs, LoopArgs
 from .deforum_controlnet import setup_controlnet_ui
+from .deforum_animatediff import setup_animatediff_ui
 from .ui_elements import get_tab_run, get_tab_keyframes, get_tab_prompts, get_tab_init, get_tab_hybrid, get_tab_output
 
 def set_arg_lists():
@@ -47,11 +48,12 @@ def setup_deforum_left_side_ui():
             tab_keyframes_params = get_tab_keyframes(d, da, dloopArgs)  # Keyframes tab
             tab_prompts_params = get_tab_prompts(da)  # Prompts tab
             tab_init_params = get_tab_init(d, da, dp)  # Init tab
+            animatediff_dict = setup_animatediff_ui()  # AnimateDiff tab
             controlnet_dict = setup_controlnet_ui()  # ControlNet tab
             tab_hybrid_params = get_tab_hybrid(da)  # Hybrid tab
             tab_output_params = get_tab_output(da, dv)  # Output tab
             # add returned gradio elements from main tabs to locals()
-            for key, value in {**tab_run_params, **tab_keyframes_params, **tab_prompts_params, **tab_init_params, **controlnet_dict, **tab_hybrid_params, **tab_output_params}.items():
+            for key, value in {**tab_run_params, **tab_keyframes_params, **tab_prompts_params, **tab_init_params, **animatediff_dict, **controlnet_dict, **tab_hybrid_params, **tab_output_params}.items():
                 locals()[key] = value
 
     # Gradio's Change functions - hiding and renaming elements based on other elements