# 5 Free AI Video Generators You Can Run Locally

- Canonical: https://mattfarmer.ai/free-ai-video-generators
- Author: [Matt Farmer](https://mattfarmer.ai/matt-farmer)
- Research checked: October 20, 2026

Compare clip length, resolution, audio and realistic GPU requirements — with official downloads, setup docs and examples.

Run these models on your own hardware without a vendor charge for every render, subject to each model’s license. Your GPU, electricity, storage and time still matter. Hosted services and GPU rentals are separate, paid choices. These are free/local models, not all permissively licensed open source.

Exactly five models are ranked. These are Matt’s editorial picks, not a measured cross-model benchmark or an Artificial Analysis leaderboard. No models were installed or run for this guide. Official examples are vendor demonstrations; hardware reports retain their community attribution. Published ranges, default clip lengths and video-file metadata are not interchangeable. Unknown hard maxima are not claimed.

## #1 MiniMax H3

Best for: Matt’s quality pick.

Beautiful generations, complex action and visual consistency, with powerful reference controls and generated stereo audio. This is my quality preference, not a claim that it wins every physics task.

- [Official models / downloads](https://huggingface.co/MiniMaxAI/MiniMax-H3)
- [ComfyUI setup guide](https://docs.comfy.org/tutorials/video/minimax/minimax-h3)
- [MiniMax community license](https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/LICENSE) — Custom MiniMax terms

### Specs and controls

- Published local range: 4–15 seconds · 24 fps. Extensions and stitched shots are separate from a first generated clip.
- Native/base: 768-pixel short edge; typical 1344 × 768.
- Later processing: The official full 2K recipe adds hosted Context-IR and Regenerate stages. It is not a completely local 2K-base pipeline.
- Generated sound: Yes · stereo. The two local-base examples below include generated stereo sound.
- Reference controls: First/last-frame control uses FL2VA. Image, video and audio references use the appropriate Ref2VA workflow; the graphs are not interchangeable.

### Hardware

Evidence: Community-tested configurations.

Specific 16GB and 12GB GPU configurations exist; there is no universal consumer minimum here.

RTX 5080 16GB: pruned INT8, ~5.17-second 960 × 544 clips with audio. RTX 4070 12GB: a smaller 4B encoder through ClipProj, a character LoRA and offloading; about 40GB system RAM observed, plus shared GPU memory. This is not the full original encoder on 12GB.

- [16GB rendered tests](https://jo-nike.github.io/h3-turbo-eval/)
- [12GB report](https://github.com/sepiablue-ai/minimax-h3-turbo-lora-benchmark/tree/main/202608_larry_vs_lightx2v)
- [Exact 12GB graph](https://github.com/sepiablue-ai/minimax-h3-turbo-lora-benchmark/blob/main/202608_larry_vs_lightx2v/workflows/workflow_baseline_20step.json)

Tradeoff: The trickiest pick to configure on constrained hardware. Offloading and changed components can make it fit, but cost time, RAM and potentially quality.

### Official examples — file metadata is not a model limit

- [Local-base text to audio/video](https://huggingface.co/MiniMaxAI/MiniMax-H3) — MiniMax; H3-Base FL2VA · local 768p. Generated stereo soundtrack. Official reproducible local-base output, not a hosted-enhanced 2K sample. Delivered file: 10.13 seconds, 1344 × 768, audio track present.
- [Omni-reference result](https://huggingface.co/MiniMaxAI/MiniMax-H3) — MiniMax; H3-Base Ref2VA · local 768p. Generated stereo soundtrack. This is the generated output, not an input reference. It uses the separate Ref2VA branch. Delivered file: 5.18 seconds, 1344 × 768, audio track present.

### Advanced notes and sources

- Comfy-Org supplies pruned INT8 ConvRot, W6A8 and FP8 diffusion choices. A separate encoder and both VAEs are still required. Checkpoint file size is not peak VRAM usage.
- The 12GB report’s inspected graph uses 576 × 832, 124 frames at 24 fps (about 5.17 seconds), Qwen3-VL 4B FP8 through ClipProj and a character LoRA. Peak system RAM is roughly 37–39GiB, dedicated VRAM about 10.6GiB, with additional shared GPU memory. It does not establish parity with the original 32B encoder or every reference/audio mode.
- A separate RTX 3090 24GB report uses pruned INT8, the original quantized 32B encoder, offloading, about 32GB RAM and 40GB swap for a reported 15-second 832 × 480 stereo-audio output. This is another specific community configuration, not a speed benchmark against the other models.
- Experimental 6/8GB reports are not a recommended H3 minimum. Changed components, reduced settings and loader/platform issues matter; capping a 16GB GPU to 8GB is not testing a physical 8GB card.
- FastH3 8-Step V2 is an optional speed derivative. Fewer steps do not automatically lower peak VRAM or preserve all reference/audio controls. Hosted H3 Max is a separate product.

- [Quantized model collection](https://huggingface.co/Comfy-Org/MiniMax-H3)
- [Quantized diffusion files](https://huggingface.co/Comfy-Org/MiniMax-H3/tree/main/diffusion_models)
- [24GB community setup](https://github.com/tonyd2wild/MiniMax-H3-Local/blob/main/docs/3090-comfyui.md)
- [FastH3 V2 speed derivative](https://huggingface.co/FastVideo/FastVideo-FastH3-8-Step-V2)

## #2 LTX-2.5

Best for: Audio, connected scenes & iteration.

Great for keeping a story flowing: generated audio, connected multi-shot scenes and first/last-frame control, with fast workflows for iteration. Distilled and quantized options offer different quality and speed tradeoffs.

- [Official models / downloads](https://huggingface.co/Lightricks/LTX-2.5)
- [ComfyUI setup guide](https://docs.comfy.org/tutorials/video/ltx/ltx-2-5)
- [LTX community license](https://github.com/Lightricks/LTX-2/blob/main/LICENSE-2_x) — Custom LTX terms

### Specs and controls

- Official local example: About 5 seconds · 121 frames / 24 fps. Longer clips are workflow-dependent; a universal local hard maximum is not verified.
- Native/base: 960 × 544 in the reference two-stage example.
- Later processing: 2× latent refinement → 1920 × 1088, about Full HD. Separate high-resolution, tiled and DFR workflows also exist.
- Generated sound: Yes. The multishot web preview below is silent; the second official example has an audio track.
- Reference controls: Connected shots, first/last-frame control and audio conditioning, depending on the selected graph. Not every combination works in every workflow.

### Hardware

Evidence: Official / reference setup.

NVIDIA 32GB+ VRAM · 32GB system RAM · 100GB free storage.

Quantization, distilled models and offloading can lower the working requirement. No universal 8GB or 12GB floor is established; encoder, frames, precision and graph all matter.

- [Hardware requirements](https://docs.ltx.io/open-source-model/getting-started/system-requirements)

Tradeoff: A small, fast preview and a polished high-resolution result are different jobs. Extra refinement stages can mean more memory and more time.

### Official examples — file metadata is not a model limit

- [Connected multishot showcase](https://ltx.io/model/ltx-2-5) — LTX / Lightricks; LTX-2.5 · official multishot showcase. This published preview is silent. The model supports generated audio. This 6.12-second, 1920 × 1080 delivery file is an example, not a native-resolution or maximum-length specification. Delivered file: 6.12 seconds, 1920 × 1080, no audio track.
- [Rainy intersection · longer multishot example](https://docs.ltx.io/open-source-model/usage-guides/prompting-guide) — LTX / Lightricks; LTX-2.5 · official prompting-guide example. Published example includes audio. About 16 seconds; a longer official example, not a universal duration limit. Delivered file: 16.09 seconds, 1920 × 1080, audio track present.

### Advanced notes and sources

- The reference two-stage example uses 121 frames at 24 fps, with 960 × 544 base and a separate 2× latent refinement to 1920 × 1088. Do not substitute 1920 × 1080 for those actual dimensions.
- Published 16/20-second examples do not establish a universal single-pass ceiling. Frame counts must follow the workflow’s constraints, such as 8k + 1; arbitrary seconds will not work in every graph.
- High-resolution/tiled generation and Diffusion Fidelity Rendering are distinct workflows. Neither ‘all native single-pass 4K’ nor ‘all high resolution is just a simple upscale’ describes the whole family.
- INT8 ConvRot, NVFP4, distillation and offloading are configuration choices, not interchangeable memory guarantees. The published reference specifies CUDA/NVIDIA; do not assume a Mac or AMD path from the same VRAM figure.

- [Official code family](https://github.com/Lightricks/LTX-2)
- [Two-stage generation](https://docs.ltx.io/open-source-model/usage-guides/two-stage-generation)
- [High-resolution / DFR workflows](https://docs.ltx.io/open-source-model/vfx-post-production/native-resolution)
- [Official example gallery](https://ltx.io/model/ltx-2-5)

## #3 Wan 2.2

Best for: Ecosystem & the smaller 5B option.

An established ecosystem of LoRAs and workflows, plus a smaller 5B model that makes local video more approachable. Choose between that TI2V model and the separate, quality-oriented A14B branches.

- [Official 5B models / downloads](https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B)
- [ComfyUI setup guide](https://docs.comfy.org/tutorials/video/wan/wan2_2)
- [Wan core license · Apache 2.0](https://github.com/Wan-Video/Wan2.2/blob/main/LICENSE.txt) — Apache 2.0 core

### Specs and controls

- Default examples: About 5 seconds. 5B: 121 frames / 24 fps. A14B: 81 frames / 16 fps. These are defaults, not proven hard caps.
- Native/base: Up to 720p; 480p / 720p depends on checkpoint and workflow.
- Later processing: An upscaled export is not native 1080p output.
- Generated sound: No · silent base. Base T2V / I2V / TI2V does not generate a soundtrack. Launch-reel music is promotional audio.
- Reference controls: TI2V-5B accepts text or an image. A14B has separate text-to-video and image-to-video checkpoints. S2V and Animate are separate workflows, not soundtrack generation in the base model.

### Hardware

Evidence: Official ComfyUI offload route.

TI2V-5B should fit 8GB VRAM with native offloading. The full A14B single-GPU reference asks for 80GB+.

That 8GB statement belongs to the smaller 5B model, not A14B. Quantization, GGUF and compatible acceleration LoRAs can help specific graphs; their quality and requirements differ.

- [8GB 5B workflow](https://docs.comfy.org/tutorials/video/wan/wan2_2)
- [Full reference setup](https://github.com/Wan-Video/Wan2.2)

Tradeoff: Lots of flexibility means checking checkpoint, LoRA and workflow compatibility. Do not substitute 5B’s memory needs for A14B’s quality or vice versa.

### Official examples — file metadata is not a model limit

- [Official family launch reel](https://huggingface.co/Wan-AI/Wan2.2-T2V-A14B) — Wan-Video / Alibaba; Wan 2.2 family · shared T2V A14B / TI2V 5B card reel. Promotional sound · not generated by the silent base model. An approximately 80-second montage, not one generation. The shared reel does not establish which checkpoint made every shot. Delivered file: 80.04 seconds, 1280 × 720, audio track present.
- [Image-to-video launch reel](https://huggingface.co/Wan-AI/Wan2.2-I2V-A14B) — Wan-Video / Alibaba; Wan 2.2 I2V A14B. Promotional soundtrack. An approximately 90-second montage, not a single clip or evidence of base-model sound generation. Delivered file: 89.75 seconds, 1280 × 720, audio track present.

### Advanced notes and sources

- The main model button opens the smaller TI2V-5B model card, not a multi-gigabyte direct download. The full text-to-video A14B and image-to-video A14B weights are separate links below.
- Fewer steps and model compression are different optimizations. Acceleration adapters must match the selected model branch and workflow.
- Both official media selections are launch montages. Their 80/90-second reel lengths do not describe a single generation, and the shared family reel does not identify every shot’s checkpoint.

- [Official repository](https://github.com/Wan-Video/Wan2.2)
- [Full T2V A14B weights](https://huggingface.co/Wan-AI/Wan2.2-T2V-A14B)
- [Full I2V A14B weights](https://huggingface.co/Wan-AI/Wan2.2-I2V-A14B)
- [ComfyUI repackaged models](https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged)

## #4 HunyuanVideo 1.5

Best for: Detailed image-to-video.

Detailed results from a relatively compact 8.3B model, especially useful for bringing still images to life. The matched input and output below show what that workflow looks like.

- [Official models / downloads](https://huggingface.co/tencent/HunyuanVideo-1.5)
- [Official setup guide](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5)
- [Tencent community license](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5/blob/main/LICENSE) — Custom Tencent terms

### Specs and controls

- Default / published benchmark: About 5 seconds by default. An official 10-second 720p benchmark also exists. A hard local maximum is not verified.
- Native/base: Native 480p / 720p.
- Later processing: 1080p uses a separate super-resolution stage.
- Generated sound: No · silent base. The selected official outputs are silent; add sound in a separate stage.
- Reference controls: Text-to-video and image-to-video. The distilled 480p I2V variant uses 8 or 12 recommended steps; that is not a blanket accelerated text-to-video mode.

### Hardware

Evidence: Official / reference setup.

14GB VRAM with model offloading. A separate community WanGP route is described as low as 6GB.

WanGP (formerly Wan2GP) is a different application/configuration. Exact settings, system RAM and platform still matter. Distillation alone does not establish a lower VRAM minimum.

- [Official requirements](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5)
- [Community WanGP route](https://github.com/deepbeepmeep/WanGP)

Tradeoff: A strong still-image animation option, but audio and 1080p delivery require other stages. These are official demonstrations, not our own runtime benchmark.

### Official examples — file metadata is not a model limit

- [Woman and parrot · image to video](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5) — Tencent Hunyuan; HunyuanVideo 1.5 · exact checkpoint / SR unspecified. Silent output. A matched official input/output pair. The still is the starting image; the video is the generated result. Delivered file: 5.04 seconds, 960 × 960, no audio track. [Matching INPUT still source](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5).
- [Soda-can crush](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5) — Tencent Hunyuan; HunyuanVideo 1.5 · exact checkpoint / SR unspecified. Silent output. An official motion/interaction example, not a controlled physics test. Delivered file: 5.04 seconds, 1280 × 720, no audio track.

### Advanced notes and sources

- The released 480p I2V step-distilled model recommends 8 or 12 steps; 4 steps adds a further quality/speed compromise. It is an I2V variant, not a universal T2V acceleration setting.
- The official model card links Wan2GP, whose repository now redirects to WanGP. Its community low-VRAM route is distinct from the documented 14GB offload reference.
- The woman-and-parrot still is the input. The matching square clip is the generated output. Exact checkpoint and super-resolution settings for that published example are unspecified; its 960 × 960 file is not a native-resolution specification.

- [Distilled 480p I2V weights](https://huggingface.co/tencent/HunyuanVideo-1.5/tree/main/transformer/480p_i2v_step_distilled)
- [WanGP (formerly Wan2GP)](https://github.com/deepbeepmeep/WanGP)

## #5 Kandinsky 6.0

Best for: Short audio-video & a smaller Lite model.

A short-form audio-video family with a smaller Lite model and an MIT-licensed core. Lite is about 3B parameters, while Pro is about 29B; they are not the same memory or quality proposition.

- [Official model collection](https://huggingface.co/kandinskylab)
- [Official quick start](https://github.com/kandinskylab/kandinsky-6/blob/main/docs/quick_start.md)
- [Kandinsky core license · MIT](https://github.com/kandinskylab/kandinsky-6/blob/main/LICENSE) — MIT core · auxiliary terms vary

### Specs and controls

- Released checkpoint target: 5 seconds · 121 frames / 24 fps. These released checkpoints target five-second clips; longer single-pass capacity is not established here.
- Native/base: SD; official model-card example 864 × 480.
- Later processing: HD / Full HD involves a separate video super-resolution (VSR) stage.
- Generated sound: Yes · synchronized audio. The family supports generated sound, but both selected published previews are silent. They are not audio demonstrations.
- Reference controls: Text-to-audio/video with synchronized sound. Released distilled checkpoints use ten steps; the exact Lite/Pro, distillation and SR settings of the gallery examples are not disclosed.

### Hardware

Evidence: Documented optimized setup.

RTX 5060 Ti 16GB preset: Pro-distill, block offloading and a quantized Qwen encoder.

This does not mean every Pro variant fits fully in 16GB. Lite is a smaller-weight option; no universal tested Lite VRAM floor is established here.

- [Exact 16GB preset](https://raw.githubusercontent.com/kandinskylab/kandinsky-6/main/kandinsky/configs/devices/rtx-5060-ti.yaml)

Tradeoff: Short clips, SD base and a newer ecosystem. Core code/model terms do not automatically cover every auxiliary encoder, component or media asset.

### Official examples — file metadata is not a model limit

- [Desert buggy](https://kandinskylab.ai/models/video/) — Kandinsky Lab; Kandinsky 6.0 family · exact variant / SR undisclosed. This published preview is silent. An official visual example, not an audio demonstration. Its 640 × 360 delivery size is not the model’s maximum native resolution. Delivered file: 5.04 seconds, 640 × 360, no audio track.
- [Girl in a glowing forest](https://kandinskylab.ai/models/video/) — Kandinsky Lab; Kandinsky 6.0 family · exact variant / SR undisclosed. This published preview is silent. The model family supports generated sound; this particular published file contains no soundtrack. Delivered file: 5.04 seconds, 640 × 360, no audio track.

### Advanced notes and sources

- The family includes Lite (~3B) and Pro (~29B), with synchronized audio, including 44kHz audio. The released five-second checkpoints use 121 frames at 24 fps; distilled versions use ten steps.
- The optimized 16GB device preset selects Pro-distill, block offloading and a quantized Qwen text encoder. It is a documented setup, not a fully resident universal minimum or proof of a Lite floor.
- The selected official gallery previews are delivered at 640 × 360 and have no audio tracks. Those file properties establish neither the base model’s maximum resolution nor its sound-generation capability. Per-example Lite/Pro, distillation and SR settings are undisclosed.

- [Official repository](https://github.com/kandinskylab/kandinsky-6)
- [Pro-distill five-second model card](https://huggingface.co/kandinskylab/Kandinsky-6.0-Pro-distill-5s-Diffusers)
- [Official example gallery](https://kandinskylab.ai/models/video/)

## What can I run?

Most evidence here is NVIDIA/CUDA. VRAM is not system RAM; Apple unified memory, AMD VRAM and CPU-only memory are not interchangeable with these numbers. Encoders, VAEs, resolution, frames, precision, offloading and RAM/swap all matter. Checkpoint size is not peak VRAM.

### Around 8GB NVIDIA VRAM

Wan 2.2 TI2V-5B with ComfyUI native offloading.

Evidence: Official offload workflow. Only the smaller 5B branch. Hunyuan also has a separate community WanGP route described as low as 6GB; settings and platform matter.

- [ComfyUI setup guide](https://docs.comfy.org/tutorials/video/wan/wan2_2)
- [Community WanGP route](https://github.com/deepbeepmeep/WanGP)

### 12GB NVIDIA VRAM

H3’s specific RTX 4070, smaller-encoder / offload configuration.

Evidence: Community-tested configuration. Qwen3-VL 4B FP8 through ClipProj, a character LoRA and about 40GB system RAM observed, plus shared GPU memory. Not the full original stack or a general guarantee.

- [12GB report](https://github.com/sepiablue-ai/minimax-h3-turbo-lora-benchmark/tree/main/202608_larry_vs_lightx2v)
- [Exact 12GB graph](https://github.com/sepiablue-ai/minimax-h3-turbo-lora-benchmark/blob/main/202608_larry_vs_lightx2v/workflows/workflow_baseline_20step.json)

### 16GB NVIDIA VRAM

H3 pruned INT8 on RTX 5080; Kandinsky’s optimized RTX 5060 Ti preset; Hunyuan’s 14GB offload reference.

Evidence: Different community / official configurations. H3 has a specific audio/video result. Kandinsky uses block offload and a quantized encoder. These are not one speed benchmark or a shared minimum.

- [16GB rendered tests](https://jo-nike.github.io/h3-turbo-eval/)
- [Exact 16GB preset](https://raw.githubusercontent.com/kandinskylab/kandinsky-6/main/kandinsky/configs/devices/rtx-5060-ti.yaml)
- [Official setup guide](https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5)

### 24GB and above NVIDIA VRAM

More H3 configurations; LTX reference starts at 32GB+; full Wan A14B single-GPU reference asks for 80GB+.

Evidence: Community result / official references. 24GB is not a blanket fit for every model. Optimized routes can differ from the reference, but a larger GPU alone is not the whole setup.

- [24GB community setup](https://github.com/tonyd2wild/MiniMax-H3-Local/blob/main/docs/3090-comfyui.md)
- [Hardware requirements](https://docs.ltx.io/open-source-model/getting-started/system-requirements)
- [Official repository](https://github.com/Wan-Video/Wan2.2)

## Glossary

- VRAM / system RAM: VRAM is memory on your graphics card. System RAM is the computer’s main memory; the two are not interchangeable.
- Offloading: Moving components between GPU and main memory to fit a smaller GPU. Usually slower, and it can need substantial RAM or swap.
- Quantization: Using lower precision to reduce memory. Quality, hardware and loader support vary.
- Distillation / fewer steps: Reducing sampling work. Faster generation does not automatically mean lower peak memory.
- LoRA: A small adapter for a subject, style or workflow-specific acceleration. It must match the model and workflow.
- I2V / T2V: Image-to-video / text-to-video: starting from an image or a written prompt.
- Native / base resolution: The generated canvas before a separate upscaler or refinement stage.
- VSR / refinement: A later stage for higher-resolution delivery, not an increase in first-clip duration.

## FAQ

### Are these AI video generators really free?

They have downloadable models you can run locally without a vendor charge for every render, subject to their licenses. Your hardware, power, storage and time are not free. Hosted subscriptions, API calls and rented GPUs are separate paid services; this guide does not promise hosted free credits.

### Do they all generate sound?

No. H3, LTX-2.5 and Kandinsky 6.0 support generated audio. Base Wan 2.2 and HunyuanVideo 1.5 are silent. Wan’s launch-reel music is promotional sound, while the selected Kandinsky previews and LTX multishot web preview are silent files despite those models’ audio capability.

### Which one has an 8GB GPU route?

The official ComfyUI guide says Wan 2.2 TI2V-5B should fit 8GB VRAM with native offloading. This does not apply to full A14B. Hunyuan has a distinct community WanGP route described as low as 6GB. Neither is a guarantee for every machine, OS, resolution or graph.

### Can MiniMax H3 run on consumer hardware?

Yes, specific community configurations have results. A 16GB RTX 5080 used pruned INT8 for roughly five-second clips with audio. A 12GB RTX 4070 report used a smaller 4B encoder through ClipProj, a character LoRA and offloading, with about 40GB system RAM observed plus shared GPU memory. That does not prove the unmodified full stack fits 12GB.

### Is five seconds a hard limit?

Not for the whole list. H3 publishes a 4–15-second local range. Several other models’ roughly five-second examples are defaults, not verified maxima; Hunyuan also publishes a 10-second benchmark, and LTX has longer workflow-dependent examples. Kandinsky’s listed released checkpoints target five-second clips. A long montage, stitched shots or continuation is not a longer first generation.

### Does a 1080p demo prove native 1080p generation?

No. Delivery files, base canvases and refinement stages are different. LTX’s reference 960 × 544 base becomes 1920 × 1088 after latent refinement. Hunyuan uses separate super-resolution for 1080p, and Kandinsky uses VSR for HD/Full HD. H3’s official full 2K recipe includes hosted stages.

### Can I use the results commercially?

Terms vary: H3, LTX and Hunyuan use custom community terms; Wan’s core uses Apache 2.0 and Kandinsky’s core uses MIT. Check the actual model license and every component you use, plus rights in inputs and outputs. Model-license labels are not a blanket permission for demonstration footage, music or trademarks. Each model section links its terms.

### Where should I start installing?

Choose a route that matches your actual GPU and system RAM, then open that model’s setup guide. The model buttons open official pages rather than starting huge checkpoint downloads. Follow the exact branch’s encoder, VAE, precision and workflow instructions; vendor terms or access gates may still apply. No finished Matt installation tutorial or one-click installer is being offered here.

## Credits and methodology

Research snapshot: October 20, 2026. Clip defaults, published ranges, refinement stages and community configurations are attributed separately. There was no controlled physics, quality or runtime benchmark. Demo rights, music and trademarks remain with their owners; source provenance does not establish a blanket reuse license. Actual model access may require vendor terms.

Start with each model’s official setup documentation. No finished Matt installation tutorial, installer, workflow ZIP or course is offered here.

- [More AI tool lists](https://mattfarmer.ai/top-ai-lists)