Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

The best open-source image and video AI tools in 2026

The best open-source image and video AI tools right now are ComfyUI, Hugging Face Diffusers and HyperFrames, out of 28 we have reviewed. Each is ranked by our rating, with its price, our verdict and who it suits.

By ReviewRamble editors · Updated Nov 18, 2026 · How we review

How we ranked: by our rating out of 5, from reviews built on each product's own pricing, documentation and policies plus independent benchmarks where they exist, each checked on the date shown in the review. We take no affiliate commission.

  1. 1

    New open image and video models land here first; the node graph and an unauthenticated server are the price of admission.

    Best for: Artists and developers who want full control over local image and video generation.

    Price: Free and open source under the GPL-3.0 licence. Comfy Cloud runs from $20 a month ($16 billed yearly); Partner Nodes are paid per use.

    Read our ComfyUI review
    4.7/5
  2. 2

    Diffusers is how open image and video models get run from code, and it assumes you will read the memory guide before you run anything large.

    Best for: Python developers and researchers building on diffusion models.

    Price: Free and open source under the Apache 2.0 license.

    Read our Hugging Face Diffusers review
    4.4/5
  3. 3

    HyperFrames makes video something a coding agent can write, and its Apache licence removes the bill that comes with Remotion.

    Best for: Developers and agent pipelines producing short product, data or documentation videos from code.

    Price: Free under Apache 2.0, with no per-render fees or commercial-use thresholds. HeyGen cloud rendering and HeyGen media (avatars, generated clips) are paid in HeyGen credits; 4K cloud renders cost 1.5 times as much.

    Read our HyperFrames review
    4.4/5
  4. 4

    AI Toolkit is where support for a new open image or video model usually lands first, and it expects you to bring the GPU and the patience.

    Best for: Creators training LoRAs for open image and video models.

    Price: Free and open source under MIT. Ostris Cloud rents preconfigured GPUs by the second, from $0.71/hour (A40 48 GB) to $2.99/hour (RTX PRO 6000 96 GB).

    Read our AI Toolkit review
    4.2/5
  5. 5

    ArcReel is the most production-minded open tool for turning stories into AI video series, provided you can live with AGPL, a single-operator design and a pipeline built around Chinese providers and Jianying.

    Best for: Small studios and solo producers making serialised AI drama, novel adaptations or product shorts who want consistent characters, reviewable stages and a clear bill per episode.

    Price: Free under AGPL-3.0; a commercial licence is available by email, price not published. You pay your image, video, text and TTS providers per call.

    Read our ArcReel review
    4.1/5
  6. 6

    MoneyPrinterTurbo is a mature, provider-agnostic way to churn out faceless short videos at almost no cost, and the output looks exactly as templated as that sounds.

    Best for: Developers and small media teams who want a self-hosted, scriptable short-video generator built from providers they already use, and will edit the output.

    Price: Free under MIT. Costs are the services you plug in: LLM tokens, optional paid TTS, and AI video clips such as Metaso MiniMax H3 at CNY 0.09 per second for 768p or CNY 0.15 for 2K. Stock footage and Edge TTS are free.

    Read our MoneyPrinterTurbo review
    4.0/5
  7. 7

    OpenFilm gets a coding agent making real videos by sticking to the web it already knows, then gives you a proper editor to finish them, though the project is only six weeks old.

    Best for: Developers and small teams who already use Claude Code or Codex and want launch videos, explainers and social clips they can still edit by hand.

    Price: Free and open source (MIT); making, editing and rendering need no account. Generated voice, music, images and video are billed by the services you connect with your own keys, or by OpenFilm account credits whose prices are not on the public site.

    Read our OpenFilm review
    4.0/5
  8. 8

    An artist-friendly local image app that has outlived its company, now mid-way through a risky v7 rewrite.

    Best for: Artists who want canvas-style image generation and editing on their own machine.

    Price: Free and open source under the Apache 2.0 licence. No paid tier since the hosted service closed in October 2025.

    Read our InvokeAI review
    3.9/5
  9. 9

    A disciplined way to make Ken Burns-style documentaries with an agent, where code-drawn camera work and a licence-checking image fetcher matter more than the four-day-old packaging.

    Best for: History and science creators, teachers and journalists who want narrated documentaries built from properly licensed archive images.

    Price: Free under MIT. Runs on your own coding agent plan; the scratch voice uses macOS say or espeak-ng at no cost, and a final voice or score costs whatever TTS or music service you choose.

    Read our Cinematic Skill review
    3.8/5
  10. 10

    Concat already does most of what people use CapCut for without an account or an upload, but it is a two-month-old beta with one main developer and the crash reports to match.

    Best for: Short-form creators and developers who want CapCut-style editing and agent automation with nothing leaving the machine.

    Price: Free and open source (AGPL-3.0). No paid tier, account or watermark; the project asks for donations on Patreon.

    Read our Concat review
    3.8/5
  11. 11

    Freefield is a thorough open-weights copy of Higgsfield's catalogue and an unusually well-documented ComfyUI project, though you need a big GPU or a Modal account and patience with model licences.

    Best for: ComfyUI users and small studios who want Higgsfield-style effects without credits and have a 24 GB GPU or a Modal account.

    Price: Free under MIT. You pay for GPU time: on Modal, a character image took 13.0 s of L40S time in the lab, about $0.007 at Modal's base price of $1.95 an hour; locally, only your own hardware.

    Read our Freefield review
    3.8/5
  12. 12

    Remocn gives a coding agent a large shelf of ready-made Remotion scenes, but its own test shows that does not automatically buy a better or cheaper video.

    Best for: Developers who already render with Remotion and want an agent to assemble product demos, launch and changelog videos from ready-made parts.

    Price: Free and MIT-licensed, with no account or usage limits. Remotion has its own licence, and larger companies may need a Remotion company licence.

    Read our Remocn review
    3.8/5
  13. 13

    Upscayl remains the simplest free way to enlarge a photo without uploading it, but the app you download today is the one from Christmas 2024.

    Best for: Photographers, designers and archivists with a discrete or recent integrated GPU who want free, private 4x enlargements of sharp but small images.

    Price: Free and open source (AGPL-3.0) from GitHub and the website; $12.99 on the Mac App Store. Upscayl Cloud offers a 10-credit free trial and Pro at $24.99 for 300 credits a month.

    Read our Upscayl review
    3.8/5
  14. 14

    A genuinely private way to turn hours of raw footage into a highlight reel, with the honest touch of showing why it picked each clip, from a small project that ships quickly.

    Best for: Hobbyists, sports clubs and travellers who shoot lots of raw footage and want highlight reels without uploading video to a cloud editor.

    Price: Free and open source under AGPL-3.0; a paid VideoHighlighter Pro and a commercial licence are offered separately.

    Read our VideoHighlighter review
    3.8/5
  15. 15

    VoiceStudio is a serious local stand-in for ElevenLabs, but the default voice model is licensed for non-commercial use and the installer asks a lot of trust.

    Best for: Individuals and small teams who want voice cloning, dubbing and transcription on their own hardware and can check each model's licence before commercial use.

    Price: Open-source app free under AGPL-3.0. Pro $99 per user per year or Lifetime $299 one-time, each for one user on up to 3 devices, with a commercial licence and batch features; Enterprise on request. Model weights carry their own licences.

    Read our VoiceStudio review
    3.8/5
  16. 16

    ArtCraft is a genuinely useful staging tool for controlling AI shots, but it is an open client for a paid credit service, not a way to run models yourself.

    Best for: Storyboard artists, previs teams and small studios who want framing control over hosted image and video models.

    Price: App is free and open source (MIT or Apache-2.0). Hosted generation uses credits: Basic $10/month (1,000 credits), Pro $35/month (3,750), Max $60/month (6,600), shown at $8, $28 and $48 billed yearly during a 20% discount; Enterprise custom. Connected Grok, Midjourney, Sora or World Labs accounts bill separately.

    Read our ArtCraft review
    3.7/5
  17. 17

    OpenMontage is the most complete attempt yet to make a coding agent behave like a production crew, though it rests on one maintainer, no releases and an AGPL licence that limits hosted use.

    Best for: Solo creators and small marketing teams already paying for a coding agent who want repeatable short videos with a visible cost trail.

    Price: Free and open source (AGPL-3.0). Zero-key path is free; paid generation runs on your own provider keys, with README examples of $1.33 to about $5 per finished short. Default budget cap $10.

    Read our OpenMontage review
    3.7/5
  18. 18

    rea-video is a sharp, unusually candid way to learn how a video was put together, as long as you read its "AI-generated" labels as informed guesses rather than detector results.

    Best for: Video producers and marketers who want to understand how a reference video was made and cost a remake, and who can treat its AI labels as a starting point.

    Price: Free under Apache-2.0. Tools-only mode runs offline at no cost; a Gemini or Doubao vision model and any video generation for remakes are billed by those providers.

    Read our rea-video review
    3.7/5
  19. 19

    A tightly opinionated reel factory: excellent if you want exactly this paper-cut style, expensive and narrow if you do not.

    Best for: A developer-relations or marketing person who wants a steady series of short paper-cut explainers about software, made with an agent they already run.

    Price: Free under MIT. The README estimates $14-19 in Claude API costs per 25-30 second reel; Remotion is free for individuals and companies of up to 3 people, paid above that.

    Read our papercut-explainer review
    3.6/5
  20. 20

    A thoughtful stills-first workflow that makes a coding agent behave like a careful motion designer, but it is a week old, has one maintainer and leans on paid tools for anything beyond motion graphics.

    Best for: Designers and small product teams who want a coding agent to storyboard and build a launch film or explainer under tight visual control.

    Price: Free under MIT. Needs Claude Code on your own plan. Optional paid extras: a clip tool such as Higgsfield for generated footage, Codex for stills, a Gemini API key for the motion check, and Suno for music (Pro from $10/month, or $8/month billed yearly; commercial use needs a paid plan).

    Read our Video Director review
    3.6/5
  21. 21

    vork is a renderer built for an agent as the artist rather than as a plug-in for a human tool, and the demos are striking, but it is an eight-day-old preview with a licence that can change and speed claims that come with long footnotes.

    Best for: Developers and motion designers who already work with coding agents and want them to produce 3D renders and VFX shots without driving a desktop editor.

    Price: Free to use, including commercially, during the research preview under a source-available licence; later versions may come under paid terms. vork cloud bills GPU time by the second from a paid Voyager workspace (Starter or above); rates are shown only on the dashboard.

    Read our vork review
    3.6/5
  22. 22

    Fuobe stacks a remarkable amount of studio on top of ComfyUI, but much of it has not yet run on a real GPU or a paid key, so it is a fast-moving preview rather than a dependable tool.

    Best for: Hobbyists and small studios with a 24 GB or larger GPU who want open-weight image, video and 3D tools without wiring ComfyUI graphs.

    Price: Free under MIT. Cloud generations are paid to fal.ai or Replicate on your own key; model weights, LoRAs and ComfyUI node packs carry their own licences.

    Read our Fuobe review
    3.5/5
  23. 23

    Image Pipes takes the guesswork out of tuning OpenCV and augmentation steps, but it is a young one-person playground with no batch mode, not a pipeline tool.

    Best for: Computer vision engineers and students who tune OpenCV preprocessing or Albumentations augmentation by eye and want the result as plain Python.

    Price: Free under MIT; no hosted or paid edition.

    Read our Image Pipes review
    3.5/5
  24. 24

    A well-engineered skill for making stickman explainers inside your coding agent, sold with a "100% free" line and a monetisation promise that neither the API bills nor YouTube's rules back up.

    Best for: Solo creators and small explainer studios who already use Codex or Claude Code and want a structured way to draft illustrated stickman videos they will edit and vary.

    Price: Free under MIT. Running it costs what your tools charge: the plan your coding agent runs on, image and voice APIs (for example Gemini 3.1 Flash Lite Image at $0.0336 per 1K image, no free tier), and a video generator for Director Mode.

    Read our Stickman Video Director review
    3.5/5
  25. 25

    A careful, local editing workflow for speech-led video that is twelve days old, tested on Macs and already stumbling on Windows.

    Best for: Mac-based creators and small teams already on Claude Code or Codex who want speech-led clips cut and captioned without uploading footage.

    Price: Free and open source (MIT). Your Claude Code or Codex subscription is separate, and any optional external generator bills you directly.

    Read our editonny review
    3.4/5
  26. 26

    SketchVLM is a clever, cheap way to make a vision model show its working on the image, worth reading and borrowing ideas from, but it is paper code without a licence and its gains are narrower than the abstract suggests.

    Best for: Researchers and product teams exploring how a vision model could show its reasoning on an image, who want a tested prompt-and-grid recipe to start from.

    Price: Code is public on GitHub with no licence file, so reuse rights are unclear. Running it costs the API calls to GPT-5, Gemini, Claude or OpenRouter models.

    Read our SketchVLM review
    3.3/5
  27. 27

    VideoAgent is an impressive paper demo and an awkward thing to run, held together by four API keys and two non-commercial model licences.

    Best for: Researchers and engineers studying how to plan multi-agent video pipelines, rather than teams that need a dependable editing tool.

    Price: Free under MIT. Running it needs paid API keys for Claude, plus DeepSeek, OpenAI and Gemini for different features; some bundled models are non-commercial.

    Read our VideoAgent review
    3.2/5
  28. 28

    OpenCut has the stars of a finished product and the code of a restart: the editor that works is archived, and the AI-ready rewrite is still a scaffold.

    Best for: People who need quick, private captions on short videos in the browser today, and developers willing to wait for the rewrite's agent hooks.

    Price: Free and open source (MIT). No pricing or paid tier is published for the hosted editor at opencut.app.

    Read our OpenCut review
    2.9/5

Sources

The main official source behind each ranked review; every review lists its full set.

More guides