Transcription
You're paying 20 bucks a month for ChatGpt Plus, and you're probably using maybe 10% of what it can actually do. I know that because I made the same mistake for almost a year. I'd open it, type a question, get an answer, close the tab. That was it.
Meanwhile, the people getting real value out of this thing were quietly using features I didn't even know existed and getting work done in minutes that was taking me hours. Welcome back to bitbiased.ai, AI, where we do the research so you don't have to. Join our community of AI enthusiasts with our free weekly newsletter. Click the link in the description below to subscribe. You will get the key AI news, tools, and learning resources to stay ahead.
So, in this video, I'm going to walk you through every model, every feature, and every tip I wish someone had told me on day one, so you can finally get your money's worth out of chat GPT. We'll cover which GPT model to actually pick for which task, the hidden tools buried inside Plus, and the prompting tricks that genuinely change the quality of what you get back.
First up, let's talk about the brain behind chat GPT. Because this is where most people get confused before they even type a single prompt, the models. A quick timeline. Here's something most people don't realize. When you open chat GPT, you're not talking to ChatGpt. You're talking to one of about a dozen different models. And which one you pick can completely change the answer you get. Let me give you the speedrun history.
Back in June 2020, OpenAI released GPT3, 175 billion parameters, API only, mostly known to developers. Then in November 2022, GPT 3.5 dropped inside the original chat GPT, and that's when the whole world noticed. March 2023 brought GPT4, which was the first one that could actually look at images. Then GPT40 in 2024 added real-time voice and video. GPT4.1 in early 2025 sharpened up coding. And then everything changed in August 2025 with GPT5.
But here's where it gets interesting. GPT5 wasn't just a smarter model. It was a system. It introduced two modes: an instant mode for quick answers and a thinking mode that actually pauses to reason through hard problems before responding. And by 2026, GPT 5.4 and GPT 5.5 push that even further. Million token context windows, way better coding, and what OpenAI calls agentic workflows, where the model can actually go do multi-step tasks on its own. So when you see those little model names in the drop down, GPT 5.4 for thinking, GPT 5.5 Pro, all of that. That's the family tree. And which one you should pick? That's the next question.
Free versus Plus versus Pro. Which models do you actually get? All right, this is the part that trips everyone up. The plan you're on directly controls which models you can talk to. If you're on the free plan, you mostly get the smaller, faster models. Think GPT5 Instant with strict daily caps. Good for quick stuff, but you'll hit a wall fast if you're doing anything heavy. The Go plan, which is OpenAI's newer interim tier, loosens those caps a bit, but still limits you to the lighter models.
Now, here's where Plus earns its $20. With Plus, you unlock GPT 5.3, GPT 5.4 Instant, and GPT 5.4 Thinking. That last one is the real upgrade. When you're stuck on a complex coding bug, a tricky negotiation email, or a strategy problem, switching to thinking mode is night and day. It literally takes longer to respond because it's working through the logic, and the answer reflects that.
Then above Plus, there's Pro, which gives you GPT 5.5 and GPT 5.5 Pro, the most advanced reasoning model OpenAI offers. That's overkill for most people. But if you're a developer, researcher, or you live inside ChatGpt all day, it's worth knowing about. Quick tip, and this saved me hours once I figured it out: Don't just leave it on auto. For anything that matters, manually switch to thinking. The quality jump is huge.
Plugins, browsing, and the Atlas browser. Now, let's get into the features, because this is where Plus really separates itself. Chat GPT can browse the live web. That sounds basic, but think about what it means. You're no longer limited to what the model learned during training. Ask it about something that happened yesterday, and it'll actually go find it. Plus, users get this enabled by default. Free users have it limited.
But the bigger story here is the GPT store. There are thousands of custom GPTs built by other users. Specialized assistance for SEO, legal research, fitness coaching, coding, and specific frameworks, you name it. Most people never touch this section and they're missing out. Spend 10 minutes browsing the store. Find two or three that match what you do daily. Pin them, and your workflow changes overnight.
And then there's Atlas, which is OpenAI's full browser. This one's wild. It doesn't just let chat GPT see one web page. It remembers everything you've browsed. So, you can ask things like, "Summarize the three articles I read this morning about AI regulation." And it actually does it. It's still rolling out, but it's a glimpse of where this is all heading.
Advanced data analysis, the feature that pays for Plus. Okay, this next part, if you do anything with data, this alone is worth the subscription. Advanced data analysis, which used to be called code interpreter, lets you upload basically any file: CSVs, Excel sheets, PDFs, even JSON. And ChatGpt will write Python code, run it, and give you back results. Charts, summaries, cleaned up data, forecasts, all of it.
Let me show you what that looks like in practice. Say you've got a messy sales spreadsheet with 3 years of data. You upload it and just type, "Show me monthly trends and forecast the next quarter." ChatGPT will actually run pandas under the hood, generate a chart, and walk you through what's happening. No formulas, no Python knowledge required. And here's the part that surprised me: it's not just for analysts. I've seen people use this to clean up messy contact lists, parse PDF invoices, even generate quick reports their boss thinks took them all afternoon. It's one of those tools where the limit is honestly your imagination.
Vision and voice, talking to ChatGpt like it's a person. This is the part that feels like science fiction the first time you try it. Vision means you can send chat GPT any image: a photo, a screenshot, a diagram, a handwritten note, and it'll understand it. Snap a picture of a menu in a foreign language, it translates it. Take a screenshot of a spreadsheet error, it explains what went wrong. Photograph a math problem from a textbook, it walks your kid through the solution.
Voice mode is the other half of this. You tap the microphone icon, and you're literally having a conversation. Plus, users get virtually unlimited voice time on the full GPT40 voice model. Free users get about 2 hours a day on the smaller version. And the voices themselves are good. Like, I forgot I was talking to a machine. Good. What I use it for personally? Driving. I'll be on a long drive and I'll just talk through ideas, get feedback, brainstorm content. It's like having a thoughtful friend on speaker phone who never gets tired of your half-baked thoughts.
Memory and custom instructions make it yours. Most people use chat GPT like a search engine. Fresh question every time. No context. That's a mistake, because there are two features sitting in your settings right now that fundamentally change how it works. Custom instructions is the first one. Go into settings, find personalization, and you'll see two boxes. The first asks what chat GPT should know about you. The second asks how you want it to respond. Fill these out properly: your role, your goals, the tone you want, the format you prefer, and every single response gets noticeably better. Mine says, "I'm a content creator. I want responses structured with clear sections, no fluff." The difference is massive.
Memory is the second one, and it's even more powerful. Chat GPT can remember details from past conversations, your projects, your preferences, your writing style, and bring them into new chats automatically. So instead of re-explaining your business every time, it just knows. Plus, users get full memory. Free users get a lighter version. And don't worry about privacy here. You can review every memory it's saved. Delete anything you don't want, or turn the whole thing off in settings.
Canvas, the editing mode you're probably ignoring. This next feature is one of those things where once you start using it, you can't go back. When you ask chat GPT to write something long, an essay, a script, a chunk of code, it can open something called canvas. That's a separate panel on the right side of the screen, and it turns chat GPT into a real editor. You can highlight any sentence and ask for changes. There are quick shortcuts for things like "make this shorter," "add more detail," "change the tone," or "explain this code." For coders, there's literally a "review my code" button that catches bugs and suggests improvements in line. The reason this matters: long-form work used to be painful in chat GPT. You'd get an output, find one paragraph you didn't like, and have to copy-paste, prompt around it, hope it didn't rewrite the whole thing. Canvas fixes that completely. You edit like you're in a Google doc, but with an AI sitting next to you the entire time.
Prompt engineering, the skill that multiplies everything. Here's the truth nobody really wants to hear: Chat GPT is only as good as your prompt. The same model can give you a generic, useless answer or a brilliant one. And the difference is almost entirely in how you ask. Three rules that change the game for me.
First, be specific. Don't say, "Write me a blog post." Say, "Write me a 600-word blog post for first-time home buyers in a friendly tone with three sections and a call to action at the end." The more constraints you give, the better the output gets. Counterintuitive, I know.
Second, give it a role. Start your prompt with something like, "You're an expert copywriter with 10 years of experience in finance." That single line completely shifts the quality of the response because now the model knows what perspective to write from.
Third, for anything complex, ask it to think step by step. With the newer GPT5 thinking models, this is built in. But even on faster models, adding "think through this carefully before answering" pushes it toward better reasoning. And one bonus tip: iterate. Your first prompt almost never gets you the best answer. Treat it like a conversation. "Make that shorter." "Try it more casual." "Add an example." That back and forth is where the magic happens.
Pricing is plus actually worth it? Let's talk money, because this is the question I get most. Free is well, free. You get GPT5 Instant with daily caps, basic voice, basic image generation, and limited browsing for casual users, students, occasional questions, light brainstorming. It's genuinely fine.
Plus is $20 a month. You get the bigger models including GPT 5. in thinking, way higher message limits, unlimited voice, full advanced data analysis, the canvas editor, custom GPTs, expanded memory, and priority access during high traffic times. For anyone using chat GPT for work, even a few times a week, it pays for itself fast.
Pro is the higher tier, pricier, aimed at power users and small teams who need GPT 5.5 Pro and basically unlimited everything. Worth it if you're a developer or researcher who lives in chat GPT, probably overkill for most people. My honest take: if you're using chat GPT more than once or twice a week for anything that matters, Plus is the obvious upgrade. If you're using it daily for serious work, look at Pro.
The honest pros and cons. Let me be real with you for a second, because no tool is perfect. The pros are obvious. Chat GPT is the most capable AI assistant most people have ever used. It writes, codes, analyzes, talks, sees images, runs Python, edits documents, and remembers context across conversations. New features ship constantly, and the gap between Plus and Free keeps getting wider in Plus's favor.
But the cons are real, too. It still hallucinates, meaning sometimes it states things confidently that are just wrong. Always fact-check anything that matters. Usage caps exist, even on Plus during peak times. By default, your conversations may be used to train future models, unless you opt out in data controls. And for genuinely nuanced work, sensitive writing, high-stakes decisions, creative projects with a strong personal voice, you still need a human in the loop. Treat it as the most powerful assistant you've ever had, not as an oracle. That mindset alone will save you from a lot of mistakes.
If this gave you even one tip you didn't know, do me a favor and hit that like button. It genuinely helps the channel reach more people who need this stuff. Subscribe if you want more deep dives like this on AI tools that actually move the needle. Drop a comment telling me which feature you're going to try first. I read all of them and I'm always curious what people latch on to. And if you want my exact prompt templates and the comparison table from this video, the link is in the description. Thanks for watching. I'll see you in the next.