AI · alternatives · 2026
24 alternatives to Text-to-video model from scratch (2 brothers, 2 years, 2B params)
Writeup (includes good/bad sample generations): https://www.linum.ai/field-notes/launch-linum-v2 We're Sahil and Manu, two brothers who spent the last 2 years training text-to-video models from scratch. Today we're releasing them under Apache 2.0. These are 2B param models capable of generating 2-5 seconds of footage at either 360p or 720p. In terms of model size, the closest comparison is Alibaba's Wan 2.1 1.3B. From our testing, we get significantly better motion capture and aesthetics. We're not claiming to have reached the frontier. For us, this is a stepping…
Linum is a text-to-video model developed by two brothers over two years, released under Apache 2.0. The 2-billion-parameter model generates 2-5 seconds of video at 360p or 720p resolution from text descriptions.… Below are 24 products that do a similar job, ranked by how close each is in meaning and then by launch-day votes.
- 1
- 2
- 3
Hi HN, author here. SHARP is Apple's recent single-image 3D Gaussian splatting model (https://arxiv.org/abs/2512.10685). Their reference code is PyTorch + a pretty heavy pipeline; I wanted to see if it could run in a browser with no server hop, so I exported the predictor to ONNX and ran it via onnxruntime-web with the WebGPU EP. What works: drop in an image, get a .ply you can download or preview live, all on your machine — your image never leaves the tab. The model is large (~2.4 GB sidecar) so first load is slow on a cold cache, but inference itself is a few seconds on…
May 2026 · github.com · its alternatives →
- 4
LTX.bio▲164Hey HN, after our 13B release we've been working on a faster version of our open-source video model and we're excited to share it. We started with a 13B base model that already had competitive generation speeds (e.g. 55s for a 5s video on an H100 — faster than any other model out there). But we wanted to push it further to allow everyone to quickly iterate over their video generations. So we built a distilled version focused on speed without sacrificing temporal or spatial coherence. With the Distilled model, you can now generate 5-second 720p videos in about 9.5 seconds on an H100, and…
2021 · lightricks.com · its alternatives →
- 5
- 6
- 7LS
Hey HN, this is Lina, Andrew, and Sidney from Lemon Slice. We’ve trained a custom diffusion transformer (DiT) model that achieves video streaming at 25fps and wrapped it into a demo that allows anyone to turn a photo into a real-time, talking avatar. Here’s an example conversation from co-founder Andrew: https://www.youtube.com/watch?v=CeYp5xQMFZY. Try it for yourself at: https://lemonslice.com/live. (Btw, we used to be called Infinity AI and did a Show HN under that name last year: https://news.ycombinator.com/item?id=41467704.) Unlike existing…
2025 · its alternatives →
- 8
Hi HN! We're proud to share Hotshot, a large-scale diffusion transformer model for text-to-video generation that we built with just a 4-person team. You can try it today in beta at https://hotshot.co, with 2 free generations per day. The model generates 5 seconds of 720p video from text prompts. It excels at prompt alignment, and consistency. It also excels at generating people, animals, and nature. In blind tests with 100 users, Hotshot generations were preferred to Runway ML 60% of the time. Hotshot generations were preferred to Luma 80% of the time. Overall, users preferred…
2024 · hotshot.co · its alternatives →
- 9BV
Vision models have been gaining popularity as a replacement for traditional OCR. Especially with Gemini 2.0 becoming cost competitive with the cloud platforms. We've been continuously evaluating different models since we released the Zerox package last year (https://github.com/getomni-ai/zerox). And we wanted to put some numbers behind it. So we’re open sourcing our internal OCR benchmark + evaluation datasets. Full writeup + data explorer here: https://getomni.ai/ocr-benchmark Github: https://github.com/getomni-ai/benchmark Huggingface:…
2025 · getomni.ai · its alternatives →
- 10
Hello HN! I want to share something me and a few friends have been working on for a while now — Zeroshot, a web tool that builds image classifiers using text-image models and autolabeling. What does this mean in practice? You can put together an image classifier in about 30 seconds that’s faster and more accurate than CLIP, but that you can deploy yourself however you’d like. It’s open source, commercially licensed, and doesn’t require you to pay anyone per API call. Here's a 2 minute video that shows it off: https://www.youtube.com/watch?v=S4R1gtmM-Lo How/why does it…
2023 · usezeroshot.com · its alternatives →
- 11
Google's most cost-effective video generation model
Apr 2026 · blog.google · its alternatives →
- 12
The model has 3B active parameters. We put the code, homepage, paper and model links here: - Code: https://github.com/bytedance/Lance - Homepage: https://lance-project.github.io/ - Paper: https://arxiv.org/abs/2605.18678 - Model: https://huggingface.co/bytedance-research/Lance p.s. Lance is a research project, not a polished product. The model was trained using fewer than 128 GPUs.
May 2026 · github.com · its alternatives →
- 13
WanX 2.1 | World's Leading AI Video Generation Platform
2025 · its alternatives →
- 14
wan2video▲14I’ve been following both Veo3 and Wan 2.5 closely, and the differences are starting to feel interesting. Veo3 has been the benchmark for cinematic AI video, especially with its stability and audio-video sync. Wan 2.5, though, takes a different route. It’s built on a native multimodal setup, meaning text, images, and audio are processed together instead of stitched from separate models. That allows smoother lip-sync, more natural background sounds, and videos that don’t feel like patchwork. The workflow is quick: input text or an image, optionally add audio, and you get a preview in minutes.…
2025 · wan2video.com · its alternatives →
- 15
Text to video & image to video cinematic creation
2025 · wan25.co · its alternatives →
- 16
- 17
It's our new text-to-image model: a 9.3B single-stream diffusion transformer trained entirely from scratch. We focused heavily on controllability through structured JSON prompts, with strong text rendering, spatial awareness through bounding box guidance, and color palette control. It has the best text rendering of any open-weight model we've tested so far, and the NF4 quantized checkpoint runs on a single 24GB GPU. For more technical details and examples see our blog post: https://ideogram.ai/blog/ideogram-4.0/ We will be happy to answer any questions :)
Jun 2026 · github.com · its alternatives →
- 18
Open-Source AI Text-to-Video Generator
2025 · stepvideot2v.com · its alternatives →
- 19
- 20
I tried to train text to video and image to video models to generate 2D game animations. I used public game animation data and 3D mixamo model rendered animations to train the animation generation models. I'm open sourcing the model, training data, training code and data generation code. More technical details can be found in my blog (https://www.gavinliblog.com/posts/godmodeanimation).
2024 · github.com · its alternatives →
- 21
AI video generation with cinematic control
2025 · wan22.io · its alternatives →
- 22
Hi HN! I built a browser-based playground for Alibaba’s new Z-Image Turbo model and wanted to share it with the community. Demo: https://z-img.net/ No GPU or setup required — everything runs through a hosted API. Key features: Photorealistic image generation in a few seconds Extremely accurate Chinese text rendering (rare among most models) Simple, lightweight interface Free to try; no login needed I created this because many makers and developers want to experiment with modern image models without downloading huge weights or renting expensive GPUs. This tool aims to be a…
Nov 2025 · z-img.net · its alternatives →
- 23
WanVideo▲8Hey HN! I built WanVideo (wanvideo.tv) to make the powerful Wan 2.2 video generation model accessible to everyone. ## What is Wan 2.2? Wan 2.2 is an advanced AI video generation model that uses an optimized MoE (Mixture of Experts) architecture to create professional-quality videos. Key capabilities: - *Cinema-Grade Aesthetics*: Professional lighting, smooth camera movements, and cinematic effects without any filming equipment - *Natural Motion Dynamics*: Realistic human movements, accurate physics simulation, and believable character expressions - *Multi-Modal Input*: Generate from text…
2025 · wanvideo.tv · its alternatives →
- 24
Open-Source, 204-Frame Video Generation from Text
2025 · github.com · its alternatives →
Also compare
Ranked by how close each launch is in meaning, then by votes. Prices were read from each product’s own site when checked and can change. Refine with your own description →