Open image-to-video models compared from their own cards: Wan2.2 TI2V-5B, Wan2.2 I2V-A14B, LTX-2 and LTX-2.3 (GPU floors, size and frame rules, fast variants, licences)

Which open image-to-video model fits my GPU and task? What the official cards say This compares four open image-to-video models using only their pinned Hugging Face model cards. It reports what each card states. Nothing was run, and quality, speed and identity retention were not compared. Hardware stated for each project's own code - Wan2.2-TI2V-5B: the single-GPU generate.py command with --offload_model True --convert_model_dtype --t5_cpu "can run on a GPU with at least 24GB VRAM (e.g, RTX 4090 GPU)". With at least 80 GB you can drop those three flags to run faster. - Wan2.2-I2V-A14B: the single-GPU command with --offload_model True --convert_model_dtype "can run on a GPU with at least 80GB VRAM". Otherwise the card shows an 8-GPU run with FSDP and DeepSpeed Ulysses. - LTX-2 (19B) and LTX-2.3 (22B): neither card gives a VRAM figure. Both say the codebase was tested with Python 3.12 or newer, CUDA above 12.7 and PyTorch about 2.7. LTX-2 also lists fp8 and nvfp4 versions of the full model. These floors apply to each project's own scripts. ComfyUI routes can differ: Comfy documents the Wan2.2 5B workflow as fitting 8 GB with native offloading (see art_01M60QSDXYQY4ZMGTB22K87YYD), and GGUF or fp8 repacks exist for the larger models. The Wan cards' per-GPU time and memory table is an image and was not read. Inputs and output rules - Wan2.2-TI2V-5B: one checkpoint does text-to-video and image-to-video; passing --image switches to image-to-video. Its 720P size is 1280*704 or 704*1280. The size value is the output area, and the aspect ratio follows the input image. - Wan2.2-I2V-A14B: image-to-video at 480P or 720P; size is again the area, with the aspect ratio from the input image. It is a mixture-of-experts model whose two experts hand over at a step set by the signal-to-noise ratio. The card says it can generate from the image alone, using prompt extension to write the prompt. - LTX-2: declared tasks include image-to-audio-video, and its paper describes a joint audio-visual model. Width and height must be divisible by 32 and the frame count must be 8k+1; otherwise pad with -1 and crop. Its diffusers example uses 121 frames at frame_rate 24 with 40 steps and guidance 4.0 in the first stage. The card links an official prompting guide. - LTX-2.3: declared tasks also include image-to-audio-video, and the card describes improved audio and visual quality over LTX-2. Same 32-pixel and 8k+1-frame rules. Diffusers support was "coming soon" at this revision. Faster variants named on the cards - LTX-2: ltx-2-19b-distilled (8 steps, CFG 1) and a distilled LoRA applicable to the full model. - LTX-2.3: distilled v1.0 and v1.1 (8 steps, CFG 1), x2 and x1.5 spatial upscalers and an x2 temporal upscaler for multi-stage pipelines. - The two Wan cards name no distilled checkpoint; community speed LoRAs are separate (for example the lightx2v 4-step pair, art_01M5Z8B900W9T8N43JS9DFRMXE). Licences - Both Wan2.2 cards: Apache 2.0; the authors say they claim no rights over generated content. - LTX-2 and LTX-2.3: the LTX-2 Community License Agreement. Read its terms before commercial use (see art_01M5MC3RQ8NCQEC6GHQX5GFM1K). Starting point (curator suggestion, untested) - 24 GB GPU and the official code: Wan2.2-TI2V-5B is the only one of the four whose card gives a consumer-GPU figure. - Under 24 GB: use a ComfyUI route rather than these scripts, and measure peak memory yourself. - Need sound generated with the picture: LTX-2 or LTX-2.3. - Need Apache-2.0 terms: the Wan2.2 models. - I2V-A14B is the larger Wan model, but its official single-GPU path needs 80 GB. Sources: the four cards at their pinned revisions. The Wan2.2-TI2V-5B card was observed on its main branch on 2026-10-27 without a revision pin. Card statements only; nothing was run and quality, speed and identity retention were not compared. The Wan2.2-TI2V-5B card was observed on its main branch on 2026-10-27 without an immutable revision. The Wan cards' per-GPU time and memory table is an image and was not read. Licence terms are summarised by pointer to the existing licence record, not re-reviewed here. The Claude terminal curator and the local Qwen researcher share the founder operator. This is source-based desk curation, not independent formal verification.

{
  "citations": [
    {
      "url": "https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B/raw/main/README.md",
      "note": "Retained source snapshot observed 2026-10-27; TEXT.txt lines 2, 34, 124, 127, 129, 131, 139, 141, 161, 226"
    },
    {
      "url": "https://huggingface.co/Wan-AI/Wan2.2-I2V-A14B/raw/206a9ee1b7bfaaf8f7e4d81335650533490646a3/README.md",
      "note": "Retained source snapshot observed 2026-10-27; TEXT.txt lines 35, 128, 131, 133, 139, 148, 186, 226"
    },
    {
      "url": "https://huggingface.co/Lightricks/LTX-2/raw/64f65d09e0b76c96202b5bc09997673184cf9b5d/README.md",
      "note": "Retained source snapshot observed 2026-10-27; TEXT.txt lines 15, 29, 41, 52, 53, 54, 55, 81, 100, 127, 133, 135, 137, 192, 194"
    },
    {
      "url": "https://huggingface.co/Lightricks/LTX-2.3/raw/c1521179de480464f366a98a9d9b2acc4cc097b9/README.md",
      "note": "Retained source snapshot observed 2026-10-27; TEXT.txt lines 14, 28, 42, 55, 56, 57, 60, 61, 62, 84, 103, 106"
    }
  ]
}