Download
1 variant available
fp32 SafeTensor
retro-90s-anime_anima1-v10-a1base-000016.safetensors
Full precision, largest file • 175.12 MB
Verified: a month ago

74.7K0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9.0 1 2 3 4 5 6 7 8 9K
3K0 1 2 3 4 5 6 7 8 9K
90K0 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9K
License:
AnimaThe Anima Model is licensed by CircleStone Labs LLC. Copyright CircleStone Labs LLC. IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
Built on NVIDIA Cosmos
This LoRA recreates the iconic look and feel of late 80s and 90s anime. Designed to emulate the bold linework, vibrant color palettes, nostalgic character archetypes, and cinematic compositions of classic hand-drawn anime, this LoRA is perfect for capturing the mood of old-school OVAs, TV dramas, and VHS-era aesthetics. From magical girls and mecha pilots to brooding detectives and high school slice-of-life scenes, this model delivers that vintage anime charm.
Usage
Sampler: ER SDE, Euler a, Euler, DPM++ 2M SDE; Schedule: Normal or Beta; Settings; 30-50 steps; 4-6 CFG; Shift 3
For best results with Anima, use quality keywords like "masterpiece, best quality, score_9, score_8, newest, recent, highres, absurdres" in the positive prompt and "low quality, worst quality, score_1, score_2, score_3, score_4, early, old, jpeg artifacts, watermark, logo" in the negative prompt.
Lora Strength: A strength between 0.8 and 1.2 is recommended. 1.0 seems perfect for most generations.
Prompting: Tags seem to work best, but a hybrid of natural language and tags also produces solid results.
Note: As a style LoRA, do not expect great results if you use this LoRA with other style LoRAs or reference incompatible styles.
I post models on multiple sites, including exclusive ones on specific platforms. Follow me on X (Twitter) or Instagram for new model releases, bounties, and more. Contact me for commissions.

