From 2ae9220f18c3df8b8d5a37caaa0a1e45fcbfd1ee Mon Sep 17 00:00:00 2001
From: creeonix <creeonix@gmail.com>
Date: Tue, 26 May 2026 14:01:28 +0100
Subject: [PATCH] Add Anima Base v1.0.

---
 configs.txt                           |  1 +
 configs/anima-base-v1.0/metadata.json | 36 +++++++++++++++++++++++++++
 models.txt                            |  1 +
 models/anima-base-v1.0/metadata.json  | 25 +++++++++++++++++++
 4 files changed, 63 insertions(+)
 create mode 100644 configs/anima-base-v1.0/metadata.json
 create mode 100644 models/anima-base-v1.0/metadata.json

diff --git a/configs.txt b/configs.txt
index ce72a23..8d398c4 100644
--- a/configs.txt
+++ b/configs.txt
@@ -1,3 +1,4 @@
+anima-base-v1.0
 ernie-image-turbo-1.0
 ernie-image-base-1.0
 anima-preview-3
diff --git a/configs/anima-base-v1.0/metadata.json b/configs/anima-base-v1.0/metadata.json
new file mode 100644
index 0000000..a20eae8
--- /dev/null
+++ b/configs/anima-base-v1.0/metadata.json
@@ -0,0 +1,36 @@
+{
+  "configuration": {
+    "batchCount": 1,
+    "batchSize": 1,
+    "clipSkip": 2,
+    "controls": [],
+    "guidanceScale": 4,
+    "height": 1024,
+    "hiresFix": false,
+    "loras": [],
+    "maskBlur": 2.5,
+    "maskBlurOutset": 0,
+    "model": "anima_base_v1.0_f16.ckpt",
+    "resolutionDependentShift": true,
+    "sampler": 16,
+    "seedMode": 2,
+    "separateClipL": false,
+    "separateOpenClipG": false,
+    "sharpness": 0,
+    "steps": 30,
+    "strength": 1,
+    "teaCache": false,
+    "t5TextEncoder": true,
+    "tiledDecoding": false,
+    "tiledDiffusion": false,
+    "upscalerScaleFactor": 0,
+    "width": 1024,
+    "teaCacheThreshold": 0.3,
+    "teaCacheStart": 5,
+    "teaCacheEnd": -1,
+    "teaCacheMaxSkipSteps": 3
+  },
+  "name": "Anima Base v1.0",
+  "version" : "cosmos2.5_2b",
+  "negative": "worst quality, low quality, score_1, score_2, score_3, artist name"
+}
diff --git a/models.txt b/models.txt
index 80d3eca..92ced7f 100644
--- a/models.txt
+++ b/models.txt
@@ -1,3 +1,4 @@
+anima-base-v1.0
 anima-preview-3
 flux-2-klein-9b
 flux-2-klein-9b-kv
diff --git a/models/anima-base-v1.0/metadata.json b/models/anima-base-v1.0/metadata.json
new file mode 100644
index 0000000..104b33c
--- /dev/null
+++ b/models/anima-base-v1.0/metadata.json
@@ -0,0 +1,25 @@
+{
+  "download": {
+    "file": "https://huggingface.co/circlestone-labs/Anima/resolve/main/split_files/diffusion_models/anima-base-v1.0.safetensors"
+  },
+  "name" : "Anima Base v1.0",
+  "version" : "cosmos2.5_2b",
+  "autoencoder" : "qwen_image_vae_f16.ckpt",
+  "prefix" : "",
+  "default_scale" : 16,
+  "hires_fix_scale" : 32,
+  "clip_encoder" : "anima_base_v1.0_f16.ckpt",
+  "file" : "anima_base_v1.0_f16.ckpt",
+  "upcast_attention" : false,
+  "text_encoder" : "qwen_3_0.6b_q8p.ckpt",
+  "high_precision_autoencoder": false,
+  "objective" : {
+    "u" : {
+      "condition_scale" : 1
+    }
+  },
+  "padded_text_encoding_length": 512,
+  "note" : "[Anima](https://huggingface.co/circlestone-labs/Anima) is a 2B text-to-image model from CircleStone Labs and Comfy Org, focused on anime concepts, characters, styles, illustrations, and other non-photorealistic art (it does not do realism well). Base v1.0 is the pretrained base model: unrefined and not aesthetic-tuned, offering maximum flexibility, diversity, and style adherence. Built on NVIDIA Cosmos and released under the CircleStone Labs non-commercial license. It works at 512-1536px, with 30-50 sampling steps and CFG 4-5 recommended.",
+  "copyright" : "© 2026 CircleStone Labs",
+  "hugging_face_link": "circlestone-labs/Anima"
+}