Benable — create shareable lists of things you recommend!
E.g., products you love, local businesses, travel recs - you can add anything to a Benable list!
My Favorite AI Tools That I Use Regularly
12
items A collection of AI tools I actually use in my day-to-day work and creative projects, from coding agents and AI assistants to image, diagram, and detection tools. Regularly updated as I discover new favorites.
Sections
4
Coding
Content Creation
AI Assistants
AI Detection
Coding
AI tools I use to code faster, experiment with different models, and improve my development workflow.
OpenCode is my favorite AI coding agents right now. It supports all the great features that you'd expect in a harness: Custom agents/Sub-agents, Skills, LSP, MCP, and the flexibility to choose the model and provider you want.
It is open source, terminal-first, even though I prefer using the desktop app, which has a great UI/UX btw.
This is currently the best coding harness in my opinion and I totally recommend giving it a try.
It is open source, terminal-first, even though I prefer using the desktop app, which has a great UI/UX btw.
This is currently the best coding harness in my opinion and I totally recommend giving it a try.
OpenCode
If you are using, or you want to try OpenCode, then I'd recommend getting the OpenCode Go plan as well.
It is one of the best-value subscriptions I've come across. For $10/month, you get access to a rotating selection of strong open coding models (GLM 5.3, Kimi K3, DeepSeek V4 Pro, etc.), all through the same OpenCode workflow. Some of the cheaper mode...
However, OpenCode Go is only meant for small projects. Codebases with more than 20k lines of code can burn through the usage quickly.
If you're already using OpenCode, I'd definitely recommend giving Go a try.
It is one of the best-value subscriptions I've come across. For $10/month, you get access to a rotating selection of strong open coding models (GLM 5.3, Kimi K3, DeepSeek V4 Pro, etc.), all through the same OpenCode workflow. Some of the cheaper mode...
However, OpenCode Go is only meant for small projects. Codebases with more than 20k lines of code can burn through the usage quickly.
If you're already using OpenCode, I'd definitely recommend giving Go a try.
OpenCode Go | Low cost coding models for everyone
Neuralwatt is an LLM inference service provider that I use for my Pay-as-you-go needs. Using their “Energy-Based Pricing” option, which is my recommendation, you pay for the actual energy consumed by the GPU, Instead of paying for inference based on ...
That makes the pricing surprisingly interesting when you're using efficient open models, especially MoE models.
A model that does the same amount of useful work with less compute can simply cost less, rather than being charged the same token rate as a much more expensive model.
Based on my own usage, the it is cost efficient only with MoE models, but with dense models, the cost is roughly the same.
So, it's definitely worth trying if you like experimenting with open models and alternative AI infrastructure.
That makes the pricing surprisingly interesting when you're using efficient open models, especially MoE models.
A model that does the same amount of useful work with less compute can simply cost less, rather than being charged the same token rate as a much more expensive model.
Based on my own usage, the it is cost efficient only with MoE models, but with dense models, the cost is roughly the same.
So, it's definitely worth trying if you like experimenting with open models and alternative AI infrastructure.
Neuralwatt: Energy-Based Pricing Plan
Content Creation
My go-to AI tools for creating visuals, diagrams, and other content for my projects and writing.
Dreamina has become one of my go-to tools when I need to create or experiment with visuals quickly.
I use it mainly for text-to-image generation but it goes beyond that. You can generate images, edit specific parts of them, expand or upscale them, and even turn images into videos, all in the same place.
I also really like having access to multiple models from one interface. It makes it easy to experiment with different models and visual styles without jumping between a bunch of different tools.
For me, it's particularly useful for blog visuals, thumbnails, design ideas, and those moments when I have a very specific visual in mind and just want to see what it looks like.
The free credits also make it easy to experiment before committing to a subscription.
I use it mainly for text-to-image generation but it goes beyond that. You can generate images, edit specific parts of them, expand or upscale them, and even turn images into videos, all in the same place.
I also really like having access to multiple models from one interface. It makes it easy to experiment with different models and visual styles without jumping between a bunch of different tools.
For me, it's particularly useful for blog visuals, thumbnails, design ideas, and those moments when I have a very specific visual in mind and just want to see what it looks like.
The free credits also make it easy to experiment before committing to a subscription.
Dreamina
I mostly use Meta AI for one thing: generating images. And honestly, I keep coming back to it because its images have a particular feel that I don't always get from Dreamina, which is my main image generation tool.
It's hard to describe exactly what it is, but the outputs often have their own visual character. Sometimes I actually prefer that slightly different aesthetic over a more polished or technically impressive result from other models.
I wouldn't necessarily use it for every image-generation task, but when I'm looking for something that feels a little different, Meta AI is one of the first tools I try.
It's also free to use, which makes experimenting with it a no-brainer.
It's hard to describe exactly what it is, but the outputs often have their own visual character. Sometimes I actually prefer that slightly different aesthetic over a more polished or technically impressive result from other models.
I wouldn't necessarily use it for every image-generation task, but when I'm looking for something that feels a little different, Meta AI is one of the first tools I try.
It's also free to use, which makes experimenting with it a no-brainer.
Meta AI
Napkin is definitely my favorite tool for generating diagrams. It's incredibly fast, and the results look genuinely good without requiring me to spend time editing or designing everything from scratch.
What I particularly find awesome is the variety of styles and layouts. I can take the same idea and quickly try a few different visual directions until I find one that fits.
The free tier is also generous, which makes it one of those tools I can easily recommend.
What I particularly find awesome is the variety of styles and layouts. I can take the same idea and quickly try a few different visual directions until I find one that fits.
The free tier is also generous, which makes it one of those tools I can easily recommend.
Napkin AI - Turn Text into AI Diagrams and Visuals
AI Assistants
The AI assistants I regularly use for research, brainstorming, writing, learning, and everyday problem-solving.
ChatGPT is one of the AI assistants I use regularly, mostly because it has become such a well-rounded tool.
I use it for everything from brainstorming and deep searching to writing, coding, analyzing files, and generating images.
What I like is that I can throw almost any kind of task at it and usually get something useful back. It has become one of those tools that is simply always worth having around.
I use it for everything from brainstorming and deep searching to writing, coding, analyzing files, and generating images.
What I like is that I can throw almost any kind of task at it and usually get something useful back. It has become one of those tools that is simply always worth having around.
ChatGPT
Claude is another AI assistant I use regularly, and I particularly like it for writing, reasoning, and coding. Its responses often feel thoughtful and well structured, which makes it a great tool for working through complicated ideas.
The main reason I use it a little less than the others is simply the free tier's usage limits. When I have enough room to use it, though, Claude is definitely one of my go-to AI assistants.
The main reason I use it a little less than the others is simply the free tier's usage limits. When I have enough room to use it, though, Claude is definitely one of my go-to AI assistants.
Claude.ai chat
Gemini is one of the AI assistants I regularly switch to. I like having it alongside the other major AI assistants because they often approach the same problem differently.
Gemini has also become particularly interesting for research, reasoning, and working with different types of media.
Gemini has also become particularly interesting for research, reasoning, and working with different types of media.
Google Gemini
Qwen Chat is my favorite alternative to the big Western AI assistants. It is very capable, particularly for coding, reasoning, and working with long documents.
I also like using it simply because it gives me another strong model to compare against. Sometimes getting a second perspective from Qwen is enough to approach a problem completely differently.
I also like using it simply because it gives me another strong model to compare against. Sometimes getting a second perspective from Qwen is enough to approach a problem completely differently.
Qwen Chat
Gemini Notebook is one of my favorite AI tools for actually learning and researching a topic.
I especially like how it stays grounded in the sources you give it, making it much easier to explore a collection of papers, articles, videos, or documents without losing track of where the information comes from.
If you regularly deal with a lot of information, Gemini Notebook is definitely worth having in your toolkit.
I especially like how it stays grounded in the sources you give it, making it much easier to explore a collection of papers, articles, videos, or documents without losing track of where the information comes from.
If you regularly deal with a lot of information, Gemini Notebook is definitely worth having in your toolkit.
Gemini Notebook (formerly NotebookLM)
AI Detection
Tools I use to experiment with AI-generated content detection and see how well these systems actually perform.
I’m generally pretty skeptical of AI detectors, but Pangram is one of the few that I actually find useful. I mainly use it when I’m curious about how likely a piece of writing is to have been AI-generated or to compare different models.
I wouldn't treat any AI detector as definitive proof, but as a tool for experimentation and getting a second signal, Pangram is the best for sure.
I wouldn't treat any AI detector as definitive proof, but as a tool for experimentation and getting a second signal, Pangram is the best for sure.
Pangram - AI Text Detection