Avis et commentaires

Kept asking for accessibility permssions over and over and over while I was using a number of different applications but NOT BoltAI. On 1st run I deleted the universal shortcut and deliberately did not give it permissions. I can copy and paste back and forth when I need to, thanks. But it really wanted those permissions. It also asked for microphone permissions when I clicked "read aloud" to get it to read out a very short answer (single sentence, just wanted to hear what it sounded like). Why? It is going text-to-speech, so no need for microphone access. A related "Why?" is that the preferences are set to use the built in Apple Speech to Text engine ... why not use the built in Text to Speech as well? Either way, it doesn't need permission to listen to my microphone to speak. Fix the un-needed permissions grab and I'll try it again.

The concept is exactly the right approach, and there are a few small things that are still being developed and improved. But it works really well, keep up the good work - BoltAi ❤️

En general funciona muy bien, permite en un chat usar diferentes modelos y eso es muy positivo. Echo de menos la interfaz en Español, pero tampoco es un problema. Lo que más me ha gustado es que puedes mantener el flujo de trabajo y usar diferentes modelos sin estar cambiando de web en web, y poder tener todos los chats por temáticas, eso se agradece. Lo que no me ha gustado nada, es que las respuestas del chat no son en Streaming, primero las calcula y luego las suelta de golpe. Si activas el streaming aparece un error en todas las peticiones del chat, da igual el modelo, falla siempre al menos a mi. Si se puede corregir esto y eliminar algunos modelos que no funcionan 4o, entonces sería la APP definitiva. A nivel de consumo de Tokens, depende del modelo pero en general gastas mucho más que si contratas, por ejemplo, chatGPT en versión Plus. Es recomendable tener la cuenta Pro con 250 Token, que con los 10 gratis se queda muy corto y en horas te los consumes.

A big improvement over Bolt v1. I had Ollama issues previously but now all working. Also with LM Studio so plenty of flexibility for local LLMs. However, can’t seem to figure out why I add documents to projects and they seem to be completely ignored.

BoltAI is my favorite AI app. I can add or change services, but my structure stays the same. The app in version 1 helped me get things done instead of reading blog posts about the best prompt engineering. Version 2 builds on this strong foundation and adds a really useful instant chat bar and a polished UI. Updates are released almost every week. Thank you!

goooooooooooooooooooooooooooooooooooooooooood

Is there a way to import stuff from BoltAI to BoltAI 2? Why are both available?

I’m impressed with BoltAI 2. It’s ambitious. There are a lot of features already there and many hinted at in the roadmap. Once the iOS version is released alongside cross-device sync, this will be a truly compelling package.

Great, if you want to avoid model/vendor lock-in. It's a chat organizer with some scaffolding around. Could steal some ideas from the big player and make it into a Vendor-Neutral-CoWork app.

The best.

Wonder why syncing is not enabled in the setapp version

New to BoltAI 2, but appreciate that it seems overall faster, has better management of context and settings (Memory Tool, Chat Inspector) and it has cloud sync and MCP support. Setapp AI support could improve, though it just added gpt-5.5, ideal if it had other newer options or at least a little bit of competitors included (Claude & Google, perhaps).

Won’t connect to Grok. No customer support. Uninstalling.

Looks good, but I can’t point a project at a folder, only individual files, which is tedious