It might seem that AI tools are pretty pricey, and they are, especially for things like video generation or top-tier LLMs such as o3 Pro. But there are a bunch of workarounds that can let you use top AI tools free of charge!
I’ve prepared a list of services that you can use to gain free access to:
- Top-tier LLMs like Gemini 2.5 Pro and DeepSeek (US hosted)
- Leading image generation models, such as Flux Kontext
- Frontier video generation models, like the Runaway Gen-4,
The thing is that we’re living in a time of a subsidized AI, when AI companies are ready to provide top-notch AI services free of charge to train on your data and stimulate adoption.
So, let’s see how to get the most out of this!
Together Chat
The first service you can use to access free AI models is Together Chat, which is interesting for two reasons:
First, all models here are hosted in North America, so it’s an excellent and simple way to test DeepSeek and Qwen without sharing your data with China-based services.
Second, and even more exciting, it also gives you free access to Flux Kontext, which is one of the most advanced image generation models, when you need to adjust an existing image. Although there are daily usage limits, you can still easily generate several images for free.
For instance, let’s take this shot and ask it to replace the couple’s modern outfits with early-20th-century beachwear. We’ll also add a Titanic-like sheep to the sea:
For Flux Kontext, it’s crucial to prepare the prompt correctly; otherwise, you won’t get the required image. I’m going to use this prompt:
Change the couple’s modern outfits to early-20th-century beachwear— the man now wears a sleeveless wool tank top and knee-length swim trunks; the woman now wears a modest knee-length bloomer suit with short sleeves and a sash while keeping both people in the exact same position, scale, pose, and orientation.
Add a large Edwardian-era ocean liner resembling the RMS Titanic on the right-hand horizon, sized and lit realistically for the distance.
Maintain identical subject placement, camera angle, framing, perspective, lighting, beach foreground, wave shapes, and all other image details; only the clothing and the added ship should change.
Here is the resulting image – nice!
I love how the shells on the beach stayed the same:
Now, let’s try a more challenging image where maintaining the correct perspective is essential. I’m going to add a few alpinists to this image:
Again, the correct prompt is the key:
Add a roped team of five alpinists in the near-foreground snowfield(lower third of the frame), sized for realistic distance and moving left-to-right toward the main peak. They wear bright red-and-blue winter climbing gear, helmets, and packs, with subtle footprints and soft shadows that follow the existing sunlight from the right.
Preserve everything else exactly as captured: mountain shapes, snow textures, lighting, color palette, camera angle, framing, and perspective (tele-lens look, ~70 mm, f/8). Keep the photorealistic style; no other elements or positions may change.
Great! I like that they are painted into this photo while maintaining all the details of the original shot.
If you want to learn how to properly prompt AI models and get the results you want, join the upcoming group of my Become an AI-Native Photographer course.
The next group starts on July 6, and this four-week course will teach you everything about an AI-powered photography workflow:
- AI Power-Skills for Photographers
- Capture One AI-Powered Tools
- Generative AI for Photographers
- AI-Driven Marketing for Photographers
There are still $149 Early Access seats available, but once they’re gone, the price of the course will increase to the regular price of $399:
What do I get with the course?
- First, you’ll get access to 8 hours of live lessons (four two-hour sessions) and their recordings.
- Second, you will immediately receive access to my Practical AI video course. (worth $79.99)
“The Practical AI course bridges the gap between theoretical knowledge and real world application, that tackled meaningful use cases rather than examples that aren’t useful. Unlike other courses that either oversimplify or get lost in technical weeds, Practical AI struck the perfect balance while remaining accessible.”
But that’s not all, you’ll also get:
- Printable workbook with 10 ready-to-use prompt templates
- Database of 50+ AI tools for photographers
- LLM Comparison Quick-Chart
- AI workflow cheat sheets and hotkey guides
- Sample RAW pack for practice
- 12 tested prompt templates for image generation
- 10 copy formula prompts
- Website LLM-optimization checklist
- and more!
In 4 Weeks, You’ll Be Able To:
- Reduce editing time significantly while maintaining your style and standards
- Set up simple automation for routine business tasks
- Spend less time on busywork and more time creating
- Evaluate new AI tools with confidence
- Use AI to brainstorm creative concepts and plan shoots
- Use image generation in your workflow
Upon completion, receive your official AI-Native Photographer certification—a mark of your commitment to staying ahead of the curve.
What if I can’t attend a live session?
All sessions are recorded and available exclusively to your cohort. However, the live experience includes Q&A and real-time feedback that makes a significant difference.
Become an AI-Native Photographer
All right, let’s move on to the next free AI tool!
Google AI Studio
Another excellent free service is Google AI Studio. It’s pretty amazing how few people know that you can use top-tier Google LLMs for free, including the fantastic Gemini 2.5 Pro.
These free models can use Google Search, browse URLs you provide, and even execute code.
Plus, AI Studio gives you free access to tools for generating images, speech, and even video. But the free usage limits are way more strict than for LLMs.
OK, now you might have a question: What’s the gimmick?
It’s pretty simple: Google will train on your data. So it’s better not to share any sensitive or personal data in this tool. However, for non-sensitive tasks, this free service is incredibly useful!
Open Router
The next one is Open Router. It’s a gateway to most – if not all – existing LLMs, including hundreds of niche and specialized models. But we’re interested in their free LLMs. Simply type “free” into the model search bar to see all the free models available.
That’s an excellent way to test various open-source models, such as DeepSeek and Llama 4.
Again, be careful with what you upload because all these models can train on your data. By the way, there is also a simple trick to increase the limits of the free models. By default, it is set to 50 requests per day. However, if you have at least $10 in your account, your daily limit increases to 1,000 free requests.
Image & Video Generation
Finally, there are some image and video generation services that offer free credits to test their models. Since generating videos and images requires a lot of computing power, these services have strict limits. Still, it’s an easy way to become familiar with the best AI models:
Image Generation
- DreamStudio – 100 credits one-time
- Leonardo.AI – 150 tokens/day
- Ideogram – 10 prompts/day
- Playground AI – Free plan available
