Anima
Anima needs 4.4 GB of VRAM in ComfyUI, at the diffusion stage of its own template, by file sizes read 3 November 2026. The smallest card that loads it with room to work is the RTX 5070 (12 GB), and the template downloads 5.6 GB.
- Made by
- CircleStone Labs
- Licence
- CircleStone Labs Non-Commercial License
- Released
- May 2026
- Sizes read
- 3 November 2026
Largest stage from exact file sizes, with ComfyUI's own template; LoRAs are not counted. Working memory for the generation comes on top. Open it in the ComfyUI checker
Which file on your card
For each memory size, the best file for that card: the highest-precision diffusion file that loads with room to work (one that only just fits when nothing else does), and what is left for working memory while it samples. On a larger card that can be above the file ComfyUI's template loads, which the answer above is about. "Streams" means even the smallest file is larger than the card: ComfyUI still runs it from system RAM, more slowly. FP4 files go only to the Blackwell cards that run them natively.
ComfyUI's template: BF16 diffusion model, BF16 text encoder. Stages: text encoder 1.2 GB, diffusion 4.4 GB. Download 5.6 GB.
Computed 3 November 2026Check another card, or pick a file by hand, in the ComfyUI VRAM checker. Built your own workflow around it, with LoRAs or custom nodes? Drop it into the ComfyUI workflow checker.
Where to run it
On-demand prices before tax, read 19 October to 3 November 2026.
About the model
Anima is CircleStone Labs' open image model. An anime image model from CircleStone Labs, made with Comfy Org and built on NVIDIA's Cosmos-Predict2 2B. At 2 billion parameters with a 0.6-billion-parameter text encoder, it is the smallest model here by a wide margin.
2B parameters Text encoder: Qwen3 0.6B CircleStone Labs Non-Commercial License Released May 2026
Before you download
The weights, and LoRAs made from them, are non-commercial; the images you generate may be used commercially. Read the CircleStone Labs Non-Commercial License itself before using it for work.
Good at
- Anime images between 512 and 1536 pixels square.
- Several finetunes (aesthetic, turbo) at the same size, so switching costs nothing in memory.
- Cards that cannot hold any of the larger models here.
Watch out for
- The weights, and LoRAs made from them, are non-commercial; the images you make may be used commercially.
- It is an anime model; it is not built for photographs.
Every Anima file ComfyUI loads
Sizes are the exact byte counts Hugging Face reports, read 3 November 2026. Files marked "template" are the ones ComfyUI's official workflow downloads.
| Part | Precision | Size | File |
|---|---|---|---|
| Diffusion model | BF16 · template | 4.18 GB | anima-base-v1.0.safetensors |
| Text encoder | BF16 · template | 1.19 GB | qwen_3_06b_base.safetensors |
| VAE | BF16 · template | 0.25 GB | qwen_image_vae.safetensors |
What CircleStone Labs says it needs
CircleStone Labs states no VRAM figure on the model card. The closest it comes: "Anima is a 2 billion parameter text-to-image model created via a collaboration between CircleStone Labs and Comfy Org." (source).
Those figures and this page measure different things. A vendor's figure covers its own script at its default resolution, working memory included. This page counts the weights each stage holds, which is exact, and leaves the working memory as the space that remains, because it depends on your resolution and attention kernel and nobody publishes it per model. If generation runs out of memory with a file that loads, lower the resolution before stepping down a precision. ComfyUI's README says it "can run even the biggest open source models on as low as 4GB vram + 8GB ram" by streaming weights (source), so "streams" means slower, not impossible.
LoRA training for Anima
No trainer we checked publishes a VRAM figure for training Anima yet (kohya's sd-scripts and musubi-tuner, ostris' ai-toolkit, diffusion-pipe, and CircleStone Labs' own repo). When one does, it goes here.
Questions
- How much VRAM does Anima need?
- In ComfyUI's own workflow, the largest stage is the diffusion stage at 4.4 GB, and the whole workflow downloads 5.6 GB. Working memory for generation comes on top and grows with resolution. CircleStone Labs publishes no VRAM figure.
- Can I run Anima on 8 GB of VRAM?
- Yes, with the BF16 file: its largest stage, the diffusion stage, is 4.4 GB, which loads on 8 GB with 3.6 GB left for working memory.
- Can I run Anima on 12 GB of VRAM?
- Yes, with the BF16 file: its largest stage, the diffusion stage, is 4.4 GB, which loads on 12 GB with 7.6 GB left for working memory.
- Can I run Anima on 16 GB of VRAM?
- Yes, with the BF16 file: its largest stage, the diffusion stage, is 4.4 GB, which loads on 16 GB with 11.6 GB left for working memory.
- Which Anima file should I download?
- The highest precision that loads fully on your card. On a 24 GB card that is the BF16 file (4.2 GB). ComfyUI's template downloads BF16 with the BF16 text encoder. FP4 files (NVFP4) are made for Blackwell cards (RTX 50, RTX PRO 6000, DGX Spark); on older cards, use FP8 or INT8.
- Files and sizes: circlestone-labs/Anima on Hugging Face, read from the Hugging Face API. Template files: ComfyUI's Anima guide.
- Model and licence: circlestone-labs/Anima and the CircleStone Labs Non-Commercial License.
- Card memory: manufacturers' specifications, as on each GPU page. The checker covers NVIDIA cards; ComfyUI also runs on AMD and Apple silicon, where file formats and speed differ and we have not sized them.
- Fit rule: a stage loads fully under 85% of usable memory and is tight up to 95%, the same margins the LLM pages use.
- Prices: each provider's own pricing page, on the dates beside them. Comfy Cloud's answer comes from ComfyUI's own template for that variant, which uses built-in nodes only, and shows only for the variants whose template we checked.