# TryUncle vs ChatGPT for Learning DaVinci Resolve

> **Quick answer:** ChatGPT is free to try and explains concepts well, but it doesn't see your DaVinci Resolve project. Its new screen-share feature is mobile-app only and untested inside Resolve, and it gets Fusion's node logic wrong. TryUncle ($19.97/month founder rate) watches your actual screen and draws a box around the control you need, live, on macOS.

*Published by [TryUncle](https://tryuncle.com) — the on-screen assistant that teaches DaVinci Resolve on your own screen.*
*Updated 2026-07-22 · ChatGPT's Advanced Voice screen sharing and pricing tiers, DaVinci Resolve 21, and TryUncle founder pricing, first 20 seats (July 2026) · Canonical: https://tryuncle.com/learn/davinci-resolve/tryuncle-vs-chatgpt-for-learning-davinci-resolve*

You typed "TryUncle vs ChatGPT" because you already have ChatGPT open in another tab, and you want to know if paying for a second tool is actually worth it. Fair question. Most people's first move when they get stuck in DaVinci Resolve is the chat window they already pay for, not a new app they've never heard of.

Here's the honest version, not a sales pitch dressed up as a comparison. ChatGPT is genuinely good at some of what stops you in Resolve. It is structurally unable to help with the rest, not because OpenAI did anything wrong, but because most of what stops an editor mid-project is a question about a specific screen, and ChatGPT, in the form nearly everyone actually uses it, has never seen that screen.

## What are TryUncle and ChatGPT, in one sentence each?

TryUncle is the on-screen assistant for DaVinci Resolve on macOS, ask in plain words and Uncle points at the exact control on your screen. It runs alongside the Edit, Color, and Fusion pages, watching whatever project you already have open, and answers by drawing a hand-drawn box around a control or moving your cursor to it.

ChatGPT is OpenAI's general-purpose chatbot. It answers typed or spoken questions about anything, including DaVinci Resolve, using training data and, on paid tiers, live web search, but by default it has no view of your screen, your project, or your timeline. It's built to answer questions about the world. It was never built specifically for DaVinci Resolve, or for any one piece of software.

That's the entire shape of the comparison before the detail. One is a general-purpose assistant you bring your own words to. The other is a narrow, purpose-built tool that watches a specific application and reacts to what's actually there. Everything below is the honest breakdown of where each one wins, where each one gets DaVinci Resolve wrong, what each one costs, and which one actually fits the moment you're stuck in right now.

**ChatGPT answers questions about DaVinci Resolve. TryUncle answers questions about your DaVinci Resolve.** Keep that distinction in mind through the rest of this piece, because nearly every other difference, price, accuracy, what happens when you're wrong, traces back to it.

## The quick comparison, side by side

Here's the full shape before the detail, so you're not scrolling back up while reading the rest of this piece.

| | ChatGPT | TryUncle |
| --- | --- | --- |
| What it is | General-purpose AI chatbot | On-screen assistant built specifically for DaVinci Resolve |
| Sees your project? | No, by default. Screen share exists but is mobile-app only and untested inside Resolve | Yes, live, continuously while you're asking |
| Points at the control? | No, describes it in words | Yes, draws a box or moves the cursor to it |
| Platform | Any device, web, iOS, Android, Windows, Mac | macOS only |
| Price | Free tier, then $8 to $200/month depending on tier | $19.97/month founder rate, one flat plan |
| Built for | Every subject, every question | DaVinci Resolve specifically, and only DaVinci Resolve |
| Known weak spot | Fusion's node-based visual logic, documented on the Blackmagic Forum | Fairlight audio mixing, not named as covered surface |
| Best at | Declarative facts, general scripting help, brainstorming | "Where is this control, right now, on my actual screen" |

Read that last row carefully, because it's the one that actually decides which tool solves the problem you have today. If you can put your question into words and the answer doesn't depend on seeing your project, ChatGPT is fast, free to start, and genuinely capable. If the problem is that you don't know what to call the thing you're stuck on, or the fix depends on your specific node tree or your specific matte, that's a different kind of problem, and it's the one TryUncle exists to solve.

## What can ChatGPT actually do for someone learning DaVinci Resolve?

More than a lot of skeptics give it credit for, within a specific lane. ChatGPT is strong at declarative questions: what a term means, where a setting is likely to live in the interface, how a workflow is generally structured, how to write or debug a Python script against Resolve's scripting API. None of that requires it to see your project, and none of it is where it actually struggles.

Ask it "what does a color qualifier do in DaVinci Resolve" and you'll get a solid, accurate explanation, the same way you would from a manual or a course, because that's documented, stable information that doesn't depend on your specific footage. Ask it to help structure a batch-rename script for your media pool, or explain the difference between serial and parallel nodes, and it's a genuinely capable study partner, available at 2am, free to start, with no course to enroll in first.

Where ChatGPT gets more useful than a static manual is follow-up. You can ask a clarifying question, get confused by the answer, and ask it a different way, which a PDF or a fixed video can't do. That back-and-forth is real value, and it's why so many editors already have it open in a second tab before they ever hear of a dedicated tool.

**A chatbot with no view of your project is genuinely useful for questions with one correct, documented answer, and it's working blind the moment your question depends on your specific footage.** That split runs through this entire comparison. Hold onto it.

## Where does ChatGPT actually get DaVinci Resolve wrong?

Fusion, specifically, and the evidence for this isn't a hunch, it's a documented forum discussion. A thread on the [Blackmagic Forum titled "Has anyone tried ChatGPT to get help with Resolve/Fusion?"](https://forum.blackmagicdesign.com/viewtopic.php?f=21&t=172854) describes exactly this pattern. Users who asked it for specific Fusion compositing steps, like building a moving sine wave with node-based expressions, got responses that were technically flawed and didn't match how Fusion actually behaves. The same thread separately notes that ChatGPT is a genuinely strong resource for writing, optimizing, and debugging general-purpose code, just not for Fusion's more specialized node logic and its Lua-based expression system specifically.

That split makes sense once you know why. Fusion is a canvas of connected nodes you build by dragging and wiring them together. A huge share of what makes a composite work is spatial and visual, and text-only training data captures that poorly, because most of what's ever been written about Fusion online describes it in prose, not in the actual shape of a working node graph. A language model trained overwhelmingly on text will describe a plausible-sounding node setup that simply doesn't correspond to anything you could build.

Here's a worked contrast that shows exactly where the line sits. Ask ChatGPT "what does the MediaIn1.Clip[1] expression reference in a Fusion Fuse," and it can generally explain the concept accurately, because that's documented, discussed, and stable across versions, text a model can learn reliably. Ask it "build me a bounce animation using a sine wave expression on this specific Transform node," and you're asking it to reason about a spatial node graph it has never seen and structurally cannot see, and the Blackmagic Forum thread above documents exactly that kind of request going wrong.

**A model trained on text will describe a plausible-sounding Fusion node setup that doesn't correspond to anything you could actually build.** That's not a ChatGPT-specific flaw so much as what happens when you ask any general-purpose language model about a narrow, visual, node-based tool most of the internet never described in enough structural detail for a model to learn reliably.

The practical rule that falls out of this: use ChatGPT for questions with one correct, documented answer, and verify anything it tells you about Fusion's specific node structure against Blackmagic's own [Visual Effects training guide](https://www.blackmagicdesign.com/products/davinciresolve/training) before you trust it on a real project.

## Can ChatGPT actually see your screen now?

Partly, and this is the part of the comparison that's changed the most recently, so it deserves a careful, honest answer rather than the outdated "ChatGPT can't see your screen at all" line a lot of older comparisons still repeat.

OpenAI added live screen sharing to Advanced Voice Mode, letting you share your screen and hold a real-time spoken conversation while ChatGPT watches and responds to what's actually on it, similar in shape to a video call (source: [Digital Trends' coverage of the rollout](https://www.digitaltrends.com/computing/openai-adds-video-analysis-and-screen-sharing-to-advanced-voice-mode/)). That's a real capability, not a rumor, and it's a meaningfully different product than the ChatGPT most DaVinci Resolve editors have used until now.

Here's the honest catch, and it's a load-bearing one for this comparison specifically. According to [MakeUseOf's how-to guide on the feature](https://www.makeuseof.com/how-screen-share-with-chatgpt/), screen sharing requires a ChatGPT Plus subscription, currently $20 a month, and it's described as exclusively accessible through the ChatGPT mobile app, on iOS and Android, not through the Mac or Windows desktop apps most Resolve editors would actually have open alongside their timeline. You open the mobile app, tap into Advanced Voice, and share either your whole phone screen or a specific app running on your phone. Reports on the feature's real-world reliability vary too; one write-up covering hands-on testing separately noted that the broadcast sometimes drops mid-session.

There's a genuinely useful, real account of what live guidance through this feature can feel like when it works. Todd Ponsky, MD, tested it navigating a task inside PowerPoint and described the moment plainly, quoted verbatim: **"ChatGPT immediately recognized where I was in the interface and said: 'You've got the format picture pane open, perfect.'"** He summed up the shift it represents, also quoted verbatim: "We're moving from 'AI that answers questions' to 'AI that helps you do things.'" (source: [Todd Ponsky's hands-on comparison of ChatGPT screen share and Gemini Live](https://toddponsky.substack.com/p/chatgpt-screen-share-vs-gemini-live)).

That's a real, positive, first-hand account, and it's worth taking seriously rather than dismissing. It's also not a test of DaVinci Resolve. This guide could not find a single confirmed, hands-on account of ChatGPT's screen share reading DaVinci Resolve's specific interface, on a Mac, live, the way it apparently read PowerPoint's format panel on Ponsky's setup. Whether it would recognize a color qualifier's softness slider or a Fusion node the way it recognized a familiar Office panel is genuinely untested, not confirmed working, and not confirmed broken either.

There's a second, more mechanical gap worth naming plainly. Even in the best-case scenario where ChatGPT's screen share does read your Resolve window accurately, it still only talks. There's no equivalent of Uncle's hand-drawn box or flying cursor anywhere in ChatGPT's screen-share feature. It can tell you, in words, where a control is. It cannot draw on your screen to show you.

**Even in its best-documented use, ChatGPT's screen sharing tells you where to look. It doesn't show you.** That's the mechanical gap no amount of future model improvement closes on its own, because it's a product design choice, not a capability limit.

| | ChatGPT screen share | TryUncle |
| --- | --- | --- |
| Requires | Plus subscription ($20/month) | Any TryUncle subscription |
| Platform | Mobile app (iOS/Android), per every guide this post could verify | macOS desktop, running alongside Resolve |
| Confirmed working inside DaVinci Resolve? | No independent test found | Yes, that's the entire product |
| Draws on your screen? | No, describes verbally | Yes, hand-drawn box or moving cursor |
| Continuous or on-demand? | On-demand, you open voice mode and start sharing | On-demand, per question, via Talk, Check, or Type |

## What does a real ChatGPT session look like when you're stuck in Resolve?

Concrete beats abstract, so here's what actually happens, walked through by input mode, based on how ChatGPT is documented to work.

**Typing a question in a browser tab.** Say you're grading a shot and the qualifier is grabbing the wall along with your subject's skin tone. You alt-tab out of Resolve, open ChatGPT, and type something like "my color qualifier is picking up the wall behind my subject, how do I fix it." ChatGPT, with no view of your actual matte, gives you the generally correct answer: soften the edge, narrow the hue range, or add a second qualifier and combine them. That's genuinely useful advice. It's also generic, because it has never seen your specific selection, your specific lighting, or how close your wall color actually is to your subject's skin tone. You alt-tab back into Resolve and have to translate that generic advice onto your specific matte yourself, guessing at which of the three suggestions applies to what you're actually looking at.

**Using Advanced Voice on your phone, without screen share.** Say you're mid-session and don't want to stop and type. You could open the ChatGPT mobile app and talk to it by voice, the way you'd talk to Uncle. Without screen sharing on, though, you're back to describing your problem in words first, "there's a node in my color page that seems to be fighting the one before it," and getting an answer shaped by however well you described something you may not have the vocabulary for yet.

**Using Advanced Voice with screen share, on your phone.** This is the newest option, and the one most directly comparable to TryUncle on paper. You'd need Resolve running somewhere your phone's camera or a mirrored screen share could see it, which for most desktop editors means either mirroring your Mac's display to your phone or holding your phone up to your monitor, neither of which is how anyone actually edits. This is the practical wall the mobile-only limitation runs into for a desktop application like Resolve specifically: the feature exists, but the setup required to point it at your actual editing screen is awkward enough that this guide couldn't find anyone who's actually done it for DaVinci Resolve and written about the result.

**Pasting an image into the chat.** The workaround most editors already use, and it's real, it works, and it's more useful than a blind text question. You capture your node tree, paste it into ChatGPT's chat window, and ask what's wrong. It can genuinely read the image and comment on what it sees. What it can't do is stay current as you keep working. The moment you drag a node or change a setting, that capture is stale, and you're pasting a new one for your next question. It's a snapshot, not a live view, and it never updates itself the way a tool actually watching your screen does.

In every one of these scenarios, the gap is the same. You're either describing your problem in words before ChatGPT can help, or feeding it a static image that goes stale the moment you touch anything. Neither is watching your project the way a person looking over your shoulder would.

## What does the same session look like with TryUncle?

Same three stuck moments, different mechanism entirely, because TryUncle never asks you to describe your screen in words first.

**Edit page, trimming a rough cut.** You can't remember which shortcut ripples a clip without leaving a gap. You hit Shift+Option, type "how do I ripple delete," and Uncle answers out loud with the shortcut while a box draws itself around the ripple-delete icon in the toolbar. You keep both hands on the keyboard. No tab-switching, no phone, no screenshot.

**Color page, the exact qualifier problem from the ChatGPT example above.** You hit Shift+Control, the Check shortcut, to ask "am I doing this right" without even framing a question. Uncle looks at your actual current matte, sees it's grabbing the wall along with the skin tone, and draws a box around the edge-softening control specifically, on your shot, not a generic list of three things that might apply. You drag the slider, the matte tightens, and you never had to guess which of ChatGPT's three suggestions matched your actual footage.

**Fusion page, a client asked for text that slides in from the left.** You hit Shift+Fn and ask out loud, "how do I animate this text sliding in." Uncle talks you through adding a Transform node and walks you to the keyframe controls with a moving cursor, reasoning about your actual node graph rather than describing a plausible-sounding one it's never seen, the exact failure mode documented on the Blackmagic Forum above.

The pattern across all three: you ask about the thing on your screen, Uncle answers about the thing on your screen, and you never translate a generic answer onto your specific project yourself. That translation step, the one every ChatGPT interaction in the previous section required, is exactly what live pointing removes.

**The gap between a generic answer and a specific one is the gap between guessing which suggestion applies to your footage and being shown the exact control that fixes it.** That's the entire value proposition of watching a screen instead of reading a description of one, and it's the reason the mechanism, not the model's raw intelligence, is what actually decides this comparison.

## How much does each one actually cost?

ChatGPT runs a six-tier pricing structure as of this writing:

| Tier | Price | What it's for |
| --- | --- | --- |
| Free | $0/month | Basic access, limited usage, no screen sharing |
| Go | $8/month | Extended limits over Free, a lower-cost middle option |
| Plus | $20/month | Full model access, required for Advanced Voice screen sharing |
| Pro | $100/month | A mid-tier introduced between Plus and the original Pro plan |
| Pro (original) | $200/month | Highest individual usage limits |
| Business | $25/user/month | Team accounts, admin controls |
| Enterprise | Custom | Large organizations, custom terms |

Source: [OpenAI's own pricing page](https://openai.com/chatgpt/pricing/) and a [third-party 2026 pricing breakdown](https://www.finout.io/blog/openai-pricing-in-2026), since exact tier names and limits shift and are worth checking directly before you commit.

TryUncle is one flat plan, no tiers to compare:

| Plan | Price | Terms |
| --- | --- | --- |
| Founder rate | $19.97/month | First 20 subscribers only, cancel anytime, 14-day no-questions-asked refund |
| Regular rate | $34.99/month | After the founder seats sell out, cancel anytime |

Source: [TryUncle's own pricing breakdown](https://tryuncle.com/learn/davinci-resolve/tryuncle-pricing-plans).

Read those two tables side by side and the honest picture is this: ChatGPT's free tier costs nothing and answers a real share of DaVinci Resolve questions well, the declarative ones covered earlier. If screen sharing is what you actually want to test, that's locked behind the $20 Plus tier, which lands close to TryUncle's founder rate anyway, for a feature that, per the section above, isn't confirmed to work with Resolve's desktop interface at all. **Comparing ChatGPT's free tier against TryUncle's $19.97 rate is comparing two different products, not two prices for the same thing.** The fairer comparison is Plus at $20 against TryUncle's $19.97, and at that price point, TryUncle is the one built specifically to watch your Resolve screen, while ChatGPT Plus's screen share is a general-purpose feature with no DaVinci Resolve-specific testing behind it.

## Which platforms and DaVinci Resolve versions does each require?

This is where the comparison resolves itself immediately for a meaningful share of readers, so it's worth answering plainly before anything else.

| | ChatGPT | TryUncle |
| --- | --- | --- |
| Operating system | Web, iOS, Android, Windows, macOS | macOS only |
| Runs alongside Resolve on the same machine? | Yes, as a separate app or browser tab | Yes, as a companion app |
| Screen share specifically | Mobile app only (iOS/Android), per every guide this post could verify | Native to the Mac app itself |
| DaVinci Resolve edition required | Not applicable, works independent of Resolve version | Works with current Resolve releases, Free or Studio |
| Internet required | Yes, always | Yes, always, no offline mode |

If you edit on Windows, ChatGPT is available to you in full. TryUncle isn't, full stop, no roadmap to point to. That's worth being direct about rather than glossing over, since this post is published on TryUncle's own site: **a Windows editor doesn't actually have this choice to make, because only one of these two tools runs on their machine at all.**

If you're on macOS, both are genuinely on the table, and the platform question stops mattering. The decision comes back to the mechanism covered throughout this piece: whether you want a general-purpose assistant you describe your problem to, or a purpose-built one that's already looking at it.

## Is ChatGPT actually free, and is TryUncle actually not?

Both halves of that question deserve a straight answer, since "free" gets thrown around loosely on both sides of this comparison.

ChatGPT's free tier is real. You can sign up with an email address and ask it DaVinci Resolve questions at no cost, today, with no card required. That's a genuine advantage over TryUncle, which has never had a free tier and isn't marketed as one. If your question is purely declarative, "what does a node do," "how do I structure this script," ChatGPT's free tier answers it without you spending anything.

Where "free" stops applying is the exact feature this comparison is actually about. Screen sharing, the one ChatGPT capability that comes closest to what TryUncle does natively, requires the $20 Plus tier. So the honest framing is narrower than "ChatGPT is free and TryUncle isn't." It's: **ChatGPT is free for declarative questions and costs $20 a month for the screen-aware feature that's the actual point of comparison here, while TryUncle is $19.97 a month for a screen-aware feature that's confirmed to work with DaVinci Resolve specifically, because that's the only thing it was built to do.**

TryUncle, for its part, has never claimed to be free and never will be. Founder pricing is a limited allocation, not a coupon, currently $19.97 a month for the first 20 subscribers, moving to $34.99 once those seats are gone, with a 14-day no-questions-asked refund if it doesn't earn its keep in your first week (source: TryUncle's FAQ).

## Does a chatbot actually teach you DaVinci Resolve, or just answer questions about it?

There's a real difference between those two things, and it matters more than which company built the model.

Watching a correct answer typed out for you, whether by ChatGPT or a video instructor, trains recognition. You read the explanation, it makes sense, and it feels like understanding. Recall is what happens later, when you're alone with your own footage and no answer paused beside it, and you have to generate the decision yourself. A chat window only ever exercises the first one. This isn't a knock on ChatGPT specifically; it's the same mechanism behind why watching a tutorial doesn't build a skill the way finishing your own project does, covered in more depth in [our research-backed look at the best way to learn DaVinci Resolve](https://tryuncle.com/learn/davinci-resolve/best-way-to-learn-davinci-resolve).

Educational research on scaffolding, the support Lev Vygotsky's zone of proximal development framework describes, makes a specific claim relevant here: the most useful help closes the exact gap between what you can do alone and what you can't do at all, no more (source: [Wikipedia, "Zone of proximal development"](https://en.wikipedia.org/wiki/Zone_of_proximal_development)). A generic answer to a vague question oversupplies help in one direction and undersupplies it in another, it gives you three possible fixes when you needed the one that matches your actual matte, and it can't see your matte to know which one that is.

There's harder data behind the general point too. A 2019 study published in the Proceedings of the National Academy of Sciences put students through both passive instruction and active, hands-on problem solving, then measured actual test performance against how much students felt they'd learned. The active group scored higher on the real tests and rated their own learning as worse (source: [Deslauriers, McCarty, Miller, Callaghan, Kestin, PNAS, 2019](https://www.pnas.org/doi/10.1073/pnas.1821936116)). Reading a clean, correct ChatGPT answer feels like progress precisely because it's smooth and effortless. Struggling to apply a generic answer to your own specific, messy footage is what actually builds the judgment you keep.

**A correct answer you read and a correct answer you had to apply yourself, to your own specific footage, are different mental events, and only one of them is the skill you're actually trying to build.** That's true whether the correct answer came from ChatGPT, a course, or a forum post. It's also the entire argument for why watching your actual screen and pointing at the fix, rather than describing one in the abstract, is a structurally different kind of help.

None of this means ChatGPT is useless for learning. It means its help lands on the declarative side of the ledger, and the judgment side, does this look right, is this node in the correct order, still has to be built by you, on your own project, ideally with something correcting you in the moment rather than after the fact.

## Should you use ChatGPT, YouTube, Blackmagic's free training, or a course instead?

ChatGPT didn't erase the rest of the DaVinci Resolve learning landscape, and a comparison that pretends the honest alternatives don't exist isn't doing you any favors. Here's where each one actually stands, named plainly.

| Resource | Cost | What it's actually good for | What it can't do |
| --- | --- | --- | --- |
| Blackmagic's free training | $0 | Structured, accurate fundamentals, written by the people who built the software | Fixed pace, doesn't react to your specific project |
| YouTube (Casey Faris and similar channels) | Free | Deep, specific technique videos, especially Fusion compositing | Doesn't know what you're stuck on right now, made for a general audience |
| Reddit and the Blackmagic Forum | Free | Crowd-sourced troubleshooting, honest community consensus, real documented failure cases like the Fusion thread above | Slow, inconsistent, depends on someone else's time and mood |
| Udemy, Skillshare, and similar course platforms | Roughly $14 to $32/month or per-course | Broad, sequential curriculum for a beginner starting from zero | Doesn't adapt once you're past the basics, still a video you watch |
| ChatGPT | Free tier, then $8 to $200/month | Fast, free-to-start answers to declarative questions and general scripting help | No default view of your project, screen share untested inside Resolve |
| TryUncle | $19.97/month founder rate | Live, in-app answers about your actual screen, on Edit, Color, and Fusion | macOS only, doesn't touch Fairlight with confirmed precision, no free tier |

None of these compete for exactly the same job, which is the honest reason a "best AI tool to learn DaVinci Resolve" question rarely has one clean answer. Blackmagic's own guides remain the most rigorously accurate resource for structured fundamentals, precisely because a machine didn't generate them (source: [Blackmagic's training page](https://www.blackmagicdesign.com/products/davinciresolve/training)). YouTube channels like [Casey Faris's](https://www.youtube.com/channel/UCdfDjoLF5L6lLuDCkJw0P3g) have taught real technique to a large free audience for years, especially in Fusion, the exact area where ChatGPT documentedly struggles most. Reddit and the Blackmagic Forum are where a lot of editors, including the ones quoted in the Fusion thread above, actually go to compare notes and find out a tool's real limitations before they hit them personally.

**The honest way to learn DaVinci Resolve fast in 2026 stacks these resources instead of picking one.** Start with Blackmagic's free guides for vocabulary and sequence. Keep ChatGPT open for the two-second declarative question that would otherwise cost you a menu hunt. Check Reddit or the Blackmagic Forum when you want a second opinion from someone who's hit your exact problem before. And for the moment none of that covers, mid-project, stuck on your own specific screen, not wanting to pause and describe it to something that can't see it, that's the gap TryUncle was built to close.

## Where does TryUncle stand next to the other AI tools built for Resolve?

ChatGPT isn't the only AI name that shows up when you search for help with DaVinci Resolve, and naming the real category honestly matters more than treating this as a two-horse race.

| Tool | What it does | Platform | Price |
| --- | --- | --- | --- |
| ChatGPT | General chatbot, declarative Q&A, mobile screen share | Any device | Free tier, then $8-$200/month |
| Claude (chat) | Same category as ChatGPT, similar strengths and gaps | Any device | Free tier, then paid plans |
| Claude + MCP scripting | Runs commands against your project via Resolve's API | Requires Resolve Studio | Free chat, needs Studio ($295 one-time) |
| DavinciClaude | Executes edits, silence cuts, captions, from a chat box | macOS and Windows | Free tier, then $6.39-$63/month |
| CutAgent | AI agent, executes timeline edits from natural language | macOS only | 29-299 euros/month |
| Sottocut | AI agent, similar category to CutAgent | Apple Silicon Mac only | $15/mo bring-your-own-key, or $129 lifetime |
| Eddie AI | Cross-NLE assistant, bridges into Resolve via an extension | Mac and Windows | Free tier, then $25-$1,250/month |
| DaVinci Resolve Neural Engine | Built-in automation (masking, reframing, speed retiming) | Native to Resolve | Included with Resolve Studio ($295 one-time) |
| TryUncle | Watches your screen, points at controls, never edits | macOS only | $19.97/month founder rate |

Sources: [CutAgent](https://www.cutagent.ai/), [Sottocut](https://sottocut.com), [Eddie AI](https://www.heyeddie.ai/), [DavinciClaude](https://www.davinciclaude.com/en), [samuelgursky/davinci-resolve-mcp](https://github.com/samuelgursky/davinci-resolve-mcp).

Notice the split once they're side by side instead of scattered across marketing pages. ChatGPT and Claude are close cousins, general chatbots with no native view of your project. DavinciClaude, CutAgent, and Sottocut are a different category entirely, agents that execute changes on your timeline directly rather than answering questions. TryUncle is the only row built specifically to watch and point without ever touching the timeline itself. If you want the full breakdown of that entire landscape, including where each tool genuinely helps and where it breaks down, our [comparison of AI tools to learn DaVinci Resolve](https://tryuncle.com/learn/davinci-resolve/ai-tools-to-learn-davinci-resolve) covers it in depth, and our [head-to-head against CutAgent specifically](https://tryuncle.com/learn/davinci-resolve/cutagent-ai-vs-tryuncle) goes deeper on the editing-agent category ChatGPT doesn't belong to at all.

**ChatGPT and TryUncle aren't actually competing for the exact same job, even though they both get typed into the same "AI tools for DaVinci Resolve" searches.** ChatGPT is a generalist you bring a well-worded question to. TryUncle is a specialist that's already looking at your screen. The other tools in that table split the difference in their own ways, some executing edits, some scripting against the API, none of them built the way TryUncle is, to watch and point without ever taking the click away from you.

## Do you need ChatGPT or TryUncle at all if DaVinci Resolve already has AI built in?

Fair question, and it's worth asking before you pay for either one. DaVinci Resolve ships its own AI, called the Neural Engine, and it's been quietly doing real work inside the app for years. Magic Mask isolates a person or object without you drawing a single point. Smart Reframe repositions a shot for a vertical crop. Speed Warp interpolates slow motion from footage that was never shot for it. None of that requires a chatbot, a subscription to a second app, or a typed question. It's built in, and on Resolve Studio, a lot of it ships with the one-time $295 license (source: [Blackmagic's DaVinci Resolve Studio page](https://www.blackmagicdesign.com/products/davinciresolve/studio)).

But the Neural Engine solves a different problem than ChatGPT or TryUncle do, and the difference is worth being precise about. Magic Mask does the masking for you. It doesn't teach you why the mask needs refining, or where the Alpha Output setting lives, or what to do next once you've got a rough matte. It's a do-it-for-you tool, not a point-at-the-answer tool. ChatGPT and TryUncle both sit in the second category: neither one touches your timeline or executes a change on your behalf. ChatGPT tells you, in words, what a setting does. TryUncle shows you, on your own screen, where that setting lives. The Neural Engine just does the thing.

Picture a concrete case. You're masking a subject for a vignette, Magic Mask does its pass in seconds, but the edge is chewing through your subject's hair whenever they turn their head. The Neural Engine already gave you its best automatic guess, it has no further explanation to offer, because explaining isn't what it does. Ask ChatGPT and you'll get the generally correct guidance: refine the mask manually with the Add and Subtract brushes, or track through the problem frames individually. Ask TryUncle and it looks at your actual mask, on your actual clip, and points at the specific refine control for the specific frame where the edge is failing, the same live-pointing mechanism covered in the Color page example earlier in this piece.

That split matters for a beginner in a specific way. If you've never used Magic Mask before, the Neural Engine will give you a working mask on the first try, but it won't explain what "Alpha Output" means when the result looks wrong, or why your mask flickers on motion blur. That's the moment a chatbot or TryUncle becomes useful again, not as a replacement for the Neural Engine, but as the thing that helps you understand what the Neural Engine just did to your footage.

| Job | DaVinci Resolve Neural Engine | ChatGPT | TryUncle |
| --- | --- | --- | --- |
| Executes the change | Yes, automatically | No | No |
| Explains the concept | No | Yes | Yes, but shows rather than tells |
| Needs to see your project | No, works inside Resolve directly | No, unless you're using screen share | Yes, that's the entire mechanism |
| Cost | Free tools in both editions; some Studio-only, included in the $295 one-time license | Free tier, then $8-$200/month | $19.97/month founder rate |

**DaVinci Resolve's own Neural Engine automates the task. ChatGPT and TryUncle explain or point at it. Those are three different jobs, not three competing tools for the same one.** Worth knowing before you assume the AI-tool question stops at "should I pay for a chatbot." One more wrinkle: several Neural Engine features are Studio-only, meaning the free version of Resolve doesn't get them at all, while ChatGPT and TryUncle work with either edition equally, since neither cares what license you're running.

## Does either tool actually help with Fairlight audio work?

Neither ChatGPT nor TryUncle names Fairlight as a strong point, and it's worth explaining exactly why, because the reason is different for each tool.

DaVinci Resolve's own Fairlight page already carries two AI-driven audio tools that do real, automatic work: Voice Isolation, which uses machine learning to strip out anything that isn't a human voice, with a single dial running from minimal to maximum; and the Dialogue Leveler, which analyzes a track and evens out loud peaks and soft dips without you touching a compressor. Voice Isolation ships Studio-only. Dialogue Leveler is available in both Free and Studio (source: [Blackmagic Forum's Voice Isolation Mode discussion](https://forum.blackmagicdesign.com/viewtopic.php?f=21&t=201942)). Those two tools alone cover a meaningful share of what a beginner assumes they'd need outside help for.

Where that leaves the two AI assistants this post is actually about: ChatGPT can explain what those tools do and when to reach for one over the other, competently, the same declarative strength it shows everywhere else in this comparison. Ask it "why does the Dialogue Leveler exist if I already have a compressor" and you'll get a solid, accurate answer about automatic gain balancing versus manual threshold-and-ratio control. That's genuinely useful, and it costs nothing on the free tier.

What ChatGPT can't do is listen to your actual mix and tell you why your dialogue is clipping on this specific clip. It has no ears, in the same way it has no eyes by default, and a text description of an audio problem, "my voice track sounds harsh and a little distorted around the loud parts," is a worse starting point than the waveform itself.

TryUncle's honest limitation here is separate and worth stating plainly, since honesty is the entire premise of this comparison. Its stated coverage is the Edit, Color, and Fusion pages, plus delivery specs. Fairlight isn't named as confirmed, precise territory the way the other three pages are. If your stuck moment is specifically "where is this Fairlight control," TryUncle may still point at it, since it watches whatever page is open, but it isn't marketed or tested to the same standard there as it is on Color and Fusion, and this post won't pretend otherwise.

Here's a worked case that shows the gap clearly. Say you're delivering to a client who specified -16 LUFS integrated for a streaming platform, and your mix reads -12 on the loudness meter. ChatGPT can explain what LUFS means, why -16 is a common streaming target, and generally how a limiter or the Dialogue Leveler brings a mix down to spec, all useful, all declarative. It cannot look at your loudness meter and tell you which of your five tracks is pushing the average up. TryUncle, watching your actual Fairlight page, could in principle point at your meter and your mixer, the same way it points at a qualifier's edge control on the Color page, but because Fairlight isn't the page it's been tested and marketed against with the same confidence, this post won't claim that works as reliably as the Color and Fusion examples covered earlier. The honest move if you're not sure: ask, see what it points at, and judge for yourself, backed by the 14-day refund if it doesn't hold up on your specific workflow.

**For audio specifically, the honest stack is Resolve's own Voice Isolation and Dialogue Leveler first, ChatGPT for the declarative "why does this tool exist" question second, and neither AI assistant for a claim of confirmed, precise, live audio guidance.** If your project is audio-heavy, our [guide to recording a podcast in DaVinci Resolve's Fairlight page](https://tryuncle.com/learn/davinci-resolve/how-to-record-a-podcast-in-davinci-resolve-fairlight) is a more direct next step than either AI tool covered here. None of this makes Fairlight a dead end for AI help, it just means the honest answer is narrower than the Color and Fusion examples earlier in this piece, and a comparison post that claimed otherwise wouldn't be telling you the truth.

## What happens to your data and your footage with each tool?

Worth answering plainly, since both tools work by reading something private, your questions, your screenshots, or your screen, and they don't handle it identically.

ChatGPT processes whatever you type, paste, or, on Plus and above, share via screen or voice, on OpenAI's own servers, governed by OpenAI's standard privacy policy and settings for whether your conversations train future models. If you're pasting a screenshot of client footage or a project under an NDA into a public chat tool, that's worth checking against your own confidentiality terms before you do it, the same way you'd check before using any cloud AI tool on sensitive material.

TryUncle reads your screen only at the moment you actually ask a question, not continuously, through standard macOS Screen Recording and Accessibility permissions. What it captures at that moment, a screenshot, your typed or spoken question, and the labels and positions of on-screen controls, briefly passes through a handful of third-party AI providers to generate an answer, then gets automatically deleted after 30 days. In TryUncle's own words: "We don't sell your data and we don't use it to train our own models" (source: TryUncle's Privacy Policy).

**If you're cutting a wedding video or a YouTube thumbnail test, neither tool's data handling is a meaningful risk worth losing sleep over.** If you're grading footage under a studio NDA, or your employer restricts which third-party tools can touch production media, that's a real conversation to have with whoever owns the NDA before you paste a screenshot into ChatGPT or install TryUncle, not a decision to make from a comparison post alone.

## Which tool actually fits your specific situation?

The comparison table earlier in this piece covers the broad strokes. Real situations are messier, so here are several specific ones, each with the reasoning spelled out.

**You're a complete beginner who's never opened DaVinci Resolve.** Start with Blackmagic's free training, not either AI tool. Both ChatGPT and TryUncle react to something already in front of them, a typed question or an open project, and on day one there's nothing yet for either to react to.

**You're stuck on where a specific control lives, right now, mid-project.** TryUncle is built for exactly this. ChatGPT can describe roughly where a control tends to live in an NLE, in the abstract, but it has never seen your panel layout, and a vague description gets a vague answer back.

**You want to understand what a term means before you go looking for it.** ChatGPT, free tier, no reason to pay for anything yet. This is squarely declarative territory, and it's fast, free, and reliable for exactly this kind of question.

**You're working in Fusion and getting a result you don't expect.** Skip ChatGPT for the specific node-level question, per the documented Blackmagic Forum thread above, and check Blackmagic's Visual Effects guide first, or ask TryUncle directly, since it's reasoning about your actual node graph rather than describing one from training data.

**You're on Windows.** ChatGPT, full stop, since TryUncle isn't available to you at all. Lean on Blackmagic's free training and the Blackmagic Forum's community troubleshooting alongside it.

**You're grading client footage under a strict NDA.** Read the NDA literally before you paste anything into either tool. TryUncle's privacy policy names its data providers directly (source: TryUncle's Privacy Policy), which gives your legal team something concrete to check against a contract clause. Confirm the same with OpenAI's terms before sharing sensitive screenshots there.

**You already pay for ChatGPT Plus and want to try screen sharing on Resolve.** Nothing stops you from testing it, since you're already paying for the tier. Go in with the honest expectation set by this post: it's mobile-app only, it's unconfirmed against Resolve's specific interface, and it won't draw on your screen the way Uncle does even if it reads it correctly.

## Can you actually use ChatGPT and TryUncle together?

Yes, and there's a genuinely sensible workflow for it, not just a hedge to avoid picking a side.

Nothing about either product rules out running both. ChatGPT lives in a browser tab or a separate app window. TryUncle runs as a companion app alongside DaVinci Resolve. They don't share an account, don't integrate with each other, and don't conflict in any way, since one never touches what the other does.

The practical split: keep ChatGPT open for the declarative question that would otherwise cost you a menu hunt, "what's the difference between a serial and parallel node," "write me a quick Python script to batch-rename these clips." Reach for TryUncle the moment the question depends on your actual screen, a matte that's grabbing the wrong thing, a node that's out of order, a control you can't locate fast enough to keep your flow going. **Neither tool replaces the other, because they're not answering the same category of question in the first place.**

The one real cost of running both is exactly what it sounds like, two subscriptions if you're on ChatGPT Plus and TryUncle's founder rate, roughly $40 a month combined. Weigh that against how often your actual stuck moments split between "I don't know what this term means" and "I don't know where this control is on my screen." For a lot of editors, that split runs closer to the second category than the first, which is the entire reason a tool like TryUncle exists as a category separate from a general chatbot at all.

## What are ChatGPT's honest strengths for a DaVinci Resolve learner?

Fair's fair, and a comparison that only lists weaknesses isn't an honest one.

It's free to start, with no card required, and the free tier answers a real share of DaVinci Resolve questions competently. It works on every platform, including Windows, where TryUncle simply isn't an option. It's good at general scripting help, debugging a Python script against Resolve's API, and explaining concepts in plain language, patiently, without judgment, at any hour. Its new screen-sharing feature is a real, meaningful capability upgrade, even though it's unconfirmed against DaVinci Resolve specifically and currently limited to mobile. And because it's a general-purpose tool, it can help with everything around your edit too, writing a video description, brainstorming a title, drafting a client email, jobs a Resolve-specific tool was never built to touch.

## What are ChatGPT's honest limitations for this specific use case?

By default, it has no view of your DaVinci Resolve project at all, and that's the single limitation this entire comparison keeps returning to. It documentedly gets Fusion's node-based logic wrong, per the Blackmagic Forum thread covered above. Its screen-sharing feature, the closest thing it has to watching your screen, requires a $20 Plus subscription, is mobile-app only per every guide this post could verify, and has no confirmed test against DaVinci Resolve's actual interface. Even when it can see a screen, it only describes what to do in words; it has no equivalent of drawing a box or moving a cursor to show you. And its training data has a cutoff, meaning a feature added in a recent DaVinci Resolve release, like Resolve 21's new Photo page, which shipped in June 2026 (source: [PetaPixel's coverage of the release](https://petapixel.com/2026/06/03/davinci-resolve-21-officially-released-with-new-photo-editing-ai-tools-and-much-more/)), may not be reflected in its answers at all.

## What are TryUncle's honest limitations?

The same honesty applies in the other direction. TryUncle is macOS only, with no public Windows roadmap, ruling it out entirely for a meaningful share of Resolve's own user base. It needs an internet connection to function at all, with no offline mode, since the reasoning that understands your screen runs in the cloud. It doesn't touch Fairlight audio mixing with confirmed precision; its stated coverage is Edit, Color, Fusion, and delivery specs. It has no free tier, and testing it means committing to at least the first billing cycle, offset only by the 14-day refund window. It's not a certificate program, and it isn't trying to be one. And it can't do what ChatGPT does well, general brainstorming, writing help, or answering a question that has nothing to do with what's currently on your screen.

## What do you do when ChatGPT or TryUncle gets something wrong?

Both tools will be wrong sometimes. That's not a knock unique to either one, it's true of any assistant, human or otherwise, and the honest question isn't whether that happens but what you do next.

**When ChatGPT gives you a confidently wrong answer.** This shows up most in Fusion, per the Blackmagic Forum thread covered earlier, but it can happen anywhere the underlying facts changed after its training data was collected, or anywhere the question depended on seeing something it couldn't see. The fix isn't complicated: cross-check anything structural against Blackmagic's own documentation or the [Visual Effects training guide](https://www.blackmagicdesign.com/products/davinciresolve/training) before you build on it, treat a ChatGPT answer about your specific project as a hypothesis rather than a fact, and if the same question keeps producing different answers across a few tries, that's usually a sign you've hit the edge of what it actually knows rather than a random glitch.

**When ChatGPT's answer references a feature that doesn't match what you're seeing.** This usually means a training-data cutoff problem: DaVinci Resolve 21 shipped a new Photo page in June 2026 (source: PetaPixel's coverage of the release), and any menu or panel that changed around a recent release may not be reflected accurately in an older model's answer. Recheck the specific claim against Blackmagic's own "What's New" page before trusting it.

**When TryUncle draws a box around the wrong control, or doesn't recognize what's on screen.** Ask again with a more specific phrasing, the same way you'd rephrase a search query that returned the wrong result. If a particular panel or dialog consistently doesn't get recognized, that's useful feedback worth reporting, since TryUncle only knows what it's been trained to look for on Resolve's specific interface, and interfaces change with every release. If it's wrong often enough on your actual workflow that it isn't saving you time, the 14-day refund exists precisely for that judgment call, no argument required (source: TryUncle's FAQ).

**When either tool goes silent or stops responding mid-session.** Check your internet connection first. Neither tool has an offline mode, both depend on a live connection to generate an answer, and a dropped connection looks identical to a broken feature until you've ruled out the boring explanation.

**When you can't tell if the advice is right at all.** This is the hardest case, and the one no tool, including TryUncle, fully solves. If a change is destructive, deleting media, overwriting a save, flattening a node tree you can't easily undo, verify it against Blackmagic's own documentation or a second source before committing, the same caution you'd apply to advice from a stranger on a forum. Neither ChatGPT nor TryUncle replaces DaVinci Resolve's own undo history, and neither should be trusted blindly on an irreversible step.

| Symptom | Likely cause | What to do |
| --- | --- | --- |
| ChatGPT confidently describes a Fusion node setup that doesn't work | Text-only training data can't capture Fusion's spatial node logic | Cross-check against Blackmagic's training guide before building on it |
| ChatGPT references a menu or feature you can't find | Training data cutoff, or a recent Resolve release moved things | Check Blackmagic's "What's New" page for the version you're running |
| TryUncle draws a box around the wrong control | Ambiguous phrasing, or an interface element it hasn't been trained to recognize yet | Rephrase more specifically; use the 14-day refund if it's a pattern on your workflow |
| Either tool stops responding | No internet connection; neither has an offline mode | Check your connection before assuming a feature broke |
| You can't tell if advice is safe to follow | Neither tool replaces judgment on irreversible steps | Verify against Resolve's own documentation before deleting, overwriting, or flattening anything |

**Getting a wrong answer occasionally isn't a reason to write off either tool, it's a reason to keep the habit of checking anything structural or irreversible against a primary source before you act on it.** That habit costs a few seconds, and it's the difference between a tool that speeds you up and one that quietly wastes an afternoon.

## Is TryUncle worth it over just using ChatGPT for free?

Depends entirely on the shape of what's actually stopping you, and that's not a dodge, it's the honest answer this whole comparison has been building toward.

If most of your stuck moments are declarative, "what does this term mean," "how do I structure this script," ChatGPT's free tier already covers you, and paying for anything else would be solving a problem you don't have. If most of your stuck moments are "I don't know where this control is, right now, on my actual screen," or "this matte is grabbing the wrong thing and I can't tell why," that's a fundamentally different kind of question, one that depends on your specific project rather than general knowledge, and it's the one ChatGPT, even with its newest features, isn't built to answer yet, while TryUncle is built around answering nothing else.

**Guided practice on your own screen beats a generic answer to a question you had to describe in words first, when the actual goal is getting unstuck fast, not just getting an answer that sounds right.** That's the entire argument, stated as plainly as the evidence in this post supports it.

## Verdict: ChatGPT or TryUncle?

If your question is declarative, and doesn't depend on your specific footage or your specific node tree, use ChatGPT. It's free to start, fast, and genuinely good at that job. Keep it open in a tab and don't feel bad about it.

If you're mid-project, stuck on a control you can't find, a matte that's grabbing the wrong thing, or a Fusion setup that isn't behaving the way a chatbot confidently told you it should, that's the specific, narrow gap TryUncle was built to close, and no amount of screen-sharing improvement on ChatGPT's side has closed it yet, at least not for DaVinci Resolve specifically, as of this writing.

Don't take this post's word over your own test. Open a real project, get stuck on purpose, and ask both tools the same question. See which one actually gets you unstuck faster, the one describing a fix in words, or the one drawing a box around it on your own screen. That test is the only review that matters, and you can run the second half of it yourself inside TryUncle.

## FAQ

### Can ChatGPT actually teach me DaVinci Resolve?

For declarative questions, what a term means, where a setting usually lives, how to structure a script, yes, reasonably well. For anything that depends on seeing your actual project, a color qualifier grabbing the wrong tone, a Fusion node out of order, it can't help beyond a guess, because in its default text mode it has no view of your screen at all.

### Can ChatGPT see my screen while I use DaVinci Resolve?

Only through its Advanced Voice screen-share feature, which requires a Plus subscription ($20/month) and, per every how-to guide this post could verify, runs through the ChatGPT mobile app rather than the Mac desktop app most Resolve editors would actually be running it alongside. Nobody has published a confirmed test of it reading DaVinci Resolve's interface specifically.

### Is TryUncle better than ChatGPT for learning DaVinci Resolve?

They do different jobs. ChatGPT is free to start and good at explaining concepts with no view of your project. TryUncle is a paid, macOS-only assistant that watches your actual DaVinci Resolve screen and draws a box around the control you need, live. If the question is declarative, ChatGPT answers it fine. If the question is 'where is this, on my screen, right now,' TryUncle is built for that and ChatGPT currently isn't.

### How much does ChatGPT cost compared to TryUncle?

ChatGPT has a free tier, then Go at $8/month, Plus at $20/month (required for screen sharing), and Pro tiers at $100 and $200/month, plus Business and Enterprise plans. TryUncle is one flat plan: $19.97/month at the founder rate for the first 20 subscribers, $34.99/month after. Check both companies' current pricing before you commit.

### What's the best way to learn DaVinci Resolve fast?

Guided, hands-on practice on your own footage, with something correcting you the moment you get a step wrong, beats watching or reading alone. A chatbot helps with the declarative facts along the way. Neither replaces finishing your own project, but a tool that watches your actual attempt and points at the fix closes more of that gap than one that answers a typed question blind.

### Is there an app that helps you while you're using DaVinci Resolve, not just a chatbot you type into separately?

Yes. TryUncle runs alongside DaVinci Resolve on macOS and watches the app while you work, answering by voice and by drawing a box or moving a cursor to the exact control on your screen. ChatGPT, in its default and most common form, is a separate window you switch to and type or talk into, with no live view of your project unless you're on the mobile app's screen-share feature.

### Does ChatGPT get DaVinci Resolve wrong sometimes?

Yes, most documented in Fusion specifically. A thread on the Blackmagic Forum describes users getting technically flawed, non-functional answers when they asked ChatGPT for help building node-based expressions, while the same thread notes it's genuinely useful for general scripting and debugging questions that don't depend on Fusion's visual node logic.

### Can I use ChatGPT and TryUncle together?

Yes, and there's a reasonable workflow for it. Use ChatGPT for declarative questions, what a term means, how to phrase a search, general Python or Lua scripting help, and use TryUncle for the moment you're stuck on your own project and need the actual control pointed at, live, without first having to describe your screen in words to something that can't see it.

## Sources

- [TryUncle](https://tryuncle.com)
- [TryUncle FAQ](https://tryuncle.com/faq)
- [TryUncle Privacy Policy](https://tryuncle.com/privacy)
- [TryUncle Pricing Plans (this hub's own pricing breakdown)](https://tryuncle.com/learn/davinci-resolve/tryuncle-pricing-plans)
- [ChatGPT Pricing (OpenAI, official page)](https://openai.com/chatgpt/pricing/)
- [OpenAI Pricing in 2026 for Individuals, Orgs & Developers](https://www.finout.io/blog/openai-pricing-in-2026)
- [OpenAI's Advanced Voice Mode can now see your screen and analyze videos (Digital Trends)](https://www.digitaltrends.com/computing/openai-adds-video-analysis-and-screen-sharing-to-advanced-voice-mode/)
- [You Can Screen Share With ChatGPT: This Is How (MakeUseOf)](https://www.makeuseof.com/how-screen-share-with-chatgpt/)
- [ChatGPT Screen Share vs. Gemini Live: My Real-World Test of AI That Watches Your Screen, by Todd Ponsky, MD](https://toddponsky.substack.com/p/chatgpt-screen-share-vs-gemini-live)
- [Blackmagic Forum - Has anyone tried ChatGPT to get help with Resolve/Fusion?](https://forum.blackmagicdesign.com/viewtopic.php?f=21&t=172854)
- [Blackmagic Forum - Voice Isolation Mode in Fairlight](https://forum.blackmagicdesign.com/viewtopic.php?f=21&t=201942)
- [DaVinci Resolve Training (Blackmagic Design)](https://www.blackmagicdesign.com/products/davinciresolve/training)
- [DaVinci Resolve product page and Studio pricing (Blackmagic Design)](https://www.blackmagicdesign.com/products/davinciresolve)
- [DaVinci Resolve - Studio (Blackmagic Design)](https://www.blackmagicdesign.com/products/davinciresolve/studio)
- [DaVinci Resolve - What's New (Blackmagic Design)](https://www.blackmagicdesign.com/products/davinciresolve/whatsnew)
- [DaVinci Resolve 21 Officially Released With New Photo Editing, AI Tools, and Much More (PetaPixel)](https://petapixel.com/2026/06/03/davinci-resolve-21-officially-released-with-new-photo-editing-ai-tools-and-much-more/)
- [GitHub - samuelgursky/davinci-resolve-mcp](https://github.com/samuelgursky/davinci-resolve-mcp)
- [GitHub - ChaiWithJai/davinci-resolve-claude-skills, by Jai Bhagat](https://github.com/ChaiWithJai/davinci-resolve-claude-skills)
- [CutAgent (product site: features, pricing, FAQ)](https://www.cutagent.ai/)
- [Sottocut (product site: platform requirements, pricing)](https://sottocut.com)
- [Eddie AI (product site: pricing, platform support)](https://www.heyeddie.ai/)
- [PremiereCopilot: 7 Best AI Plugins for DaVinci Resolve in 2026](https://www.premierecopilot.com/en/blog/best-ai-plugins-davinci-resolve-2026)
- [DavinciClaude](https://www.davinciclaude.com/en)
- [Casey Faris - YouTube channel](https://www.youtube.com/channel/UCdfDjoLF5L6lLuDCkJw0P3g)
- [MOOC completion rate just 4%, study says (Higher Ed Dive)](https://www.highereddive.com/news/mooc-completion-rate-just-4-study-says/202425/)
- [Deslauriers, McCarty, Miller, Callaghan, Kestin: Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom (PNAS, 2019)](https://www.pnas.org/doi/10.1073/pnas.1821936116)
- [Zone of proximal development - Wikipedia](https://en.wikipedia.org/wiki/Zone_of_proximal_development)