Roadmap
Help Center
Changelog

Help me improve BoltAI

For a bug report, please include the following information:

  • Your OS version.

  • Your BoltAI app version.

  • Are you a Setapp user.

  • The AI provider & model you’re using.

  • Steps to reproduce the issue.

  • The error message

BUG: UI hangs when running mcp_add_memory_tctnzi tool

OS: macOS 26.4.1 Version: BoltAI 2.12.1 (build 68) - Setapp LLM: Anthropic Claude Sonnet 4.6 I instructed the LLM to commit something in particular to memory. It appears to have called the mem0-memory-mcp server installed. Although it’s not clear to me if that’s what Bolt tried to use, or whether it’s using the built-in Bolt memory tool. I can see in the UI what you see in attached screenshot. The conversation hang after it tried to use this memory tool. I had to manually terminate the hung prompt response. I can see that the number of memories in the Bolt Memory tool didn’t increase. So if that is the memory system it’s utilising in this chat, the hung attempt to add a memory did not get so far as actually adding a memory.

Inspired Earth 5 days ago

2
💻

BoltAI Mac v2

Google Deep Research models not working

When connecting through “Google AI” AI Providers and when selecting any Deep Research model there is an error: Error This model only supports Interactions API. I have all MCPs and plugins disabled… all other Gemini models work. https://ai.google.dev/gemini-api/docs/interactions/deep-research i guess you would need to implement an additional api endpoint for this to work?

David 5 days ago

💻

BoltAI Mac v2

Choice of local models for STT

Not having a choice in speech-to-text models for the dictation feature is very limiting. The two pre‑selected models are not great. Silences are always replaced with lots of random words. And the user doesn't have the ability to pick from a whole variety of excellent speech-to-text models now available. Look at something like Mac Whisper and the flows that exist there.

Thinkun 7 days ago

4
💻

BoltAI Mac v2

BUG: Often VTT transcription doesn’t get posted into the prompt

I am finding that more often than not, when I use VTT for writing my prompt, when I click the blue checkmark button the VTT input modal closes and nothing is inserted into the prompt. I essentially lose what I dictated. Because this happens so often (more often than not) I am now in the habit of first selecting the live transcription text that appeared in the modal, so that I can manually paste it into the prompt when the UI fails to do so. This is on v2.12.0 (build 67).

Inspired Earth 7 days ago

3
💻

BoltAI Mac v2

Built a Hermes Agent gateway plugin for BoltAI — full markdown + slash commands

Hey, I've been using BoltAI v2 with Hermes Agent (https://github.com/NousResearch/hermes-agent) and hit two annoyances with the official OpenAI-compatible gateway: Markdown gets flattened — headers, tables, code blocks all come through as plain text Hermes slash commands (/help, /status, /new, /model, etc.) aren't intercepted, so they get sent to the LLM instead of running locally and therefore they dont work properly as AI agent cant work with some directly… So I built a small hermes drop-in plugin that fixes both. It's a separate gateway adapter (subclasses the official one, doesn't fork it), runs on its own port, and lives here: https://github.com/Dworf/boltai-hermes-gateway Install: hermes plugins install Dworf/boltai-hermes-gateway --enable Then in BoltAI just follow the official Hermes setup guide (https://help.boltai.com/en/help/articles/9536908-how-to-use-hermes-agent-in-boltai) but point it at port 8643 instead of 8642. Full markdown renders correctly, slash commands work as expected, streaming is preserved. Should also work with any other OpenAI-compatible client that supports markdown + streaming (TypingMind, OpenWebUI, LibreChat, etc.) — untested though (and yeah BoltAI has no competition in this space :) ) Question for the BoltAI team / community: if there's interest, I'm happy to implement a commands discovery endpoint on the plugin side so that — if BoltAI ever wants to add it on the client side — you could get interactive slash command suggestions, autocomplete, and /-menu UX natively. Let me know if that's something worth building toward. Should be easy from the plugin side… Thanks!

David 11 days ago

2
💻

BoltAI Mac v2

Model popover fetching models for a very long time

I have a couple of models on openrouter.ai and a few via Ollama on a PC in my local network set up. The local PC is turned off most of the time. Recently this popover where you can switch the current model started fetching models for a long time (30s+) and I had to close and re-open BoltAI 2 to work around it. My internet access at that time is working as usual, I can also open my openrouter.ai profile without any problems. Let me know how I can submit any logs to help debug this.

zuKoma091 11 days ago

1
💻

BoltAI Mac v2

model selection of model capability based on output (text, image, video,...)

I really like the new model selection… and love the favorites. but when working with model collections like (image generation, video generation,…) it would be good to have filter option to only show models that can do image generation or video generation,…. i would be happy to just tag each model if it supports for example image generation… i think you could pull this data as well from openrouter for example…

David 18 days ago

💻

BoltAI Mac v2

Planned

Searchable Tools activation list

I’ve recently moved from BoltAI 1.x to BoltAI 2.x. I see the tools activation has moved to an icon in the prompt field, rather than a list in the side panel. I think it’s good that it moved to an icon. However, with many MCP servers installed, and some with a massive number of tools, it’s now a very slow process to turn on any given MCP server for a chat session. It can take me minutes just to activate the tools I want to use in the chat. I suggest adding a live search tool. So the user can start typing into the search field the name of an MCP server and or tool, and the list hones in on the results, live / dynamically. Much like macOS and iOS with its settings UI. I would also suggest having a little switch on that search system to include or not the actual tools in the results, or just search the servers. In my usage, I will almost always be just searching for servers, and turning on the entire server rather than individual tools. But I guess some people want to turn on individual tools, so it might be handy for those people to be able to also search by tool name.

Inspired Earth 18 days ago

1
💻

BoltAI Mac v2

Config and data controls for dictations: (auto) delete past dictations and set custom timeout for cleanup stage

The recent improvements to dictations (which added the cleanup stage and tool bar dictation history) are great. A few suggested improvements on usage and data controls: 1) My dictations may be longer than most people (5-10 minutes sometimes). As such, the timeout on the cleanup stage frequently kicks in. Option to manually set it would be good. 2) A setting to auto delete all previous dictations after a set period (eg 24 hours) and/or a global delete button for old dictations and/or option to delete each recording from menu bar. This should also delete any audio files (if they haven’t been wiped already).

Matt 24 days ago

1
💻

BoltAI Mac v2

Support for Tailscale-gated endpoints

OS version: Tahoe 26.4.1 App version: 2.9.2 Not a Setapp user AI and provider: haven't gotten that far (see below) Steps to reproduce: attempt to add a custom provider with just a base URL and no API key → no models are detected Tailscale released their AI gateway, Aperture (still in alpha), which authenticates based on Tailscale device identity instead of an API key. I’m not sure if BoltAI rejects the lack of API key or just isn’t compatible with sending requests from my device on my Tailnet, but I figured I’d ask for support of this pattern since Tailscale’s play has a good chance of being popular. Props on the heroic effort of supporting all these weird implementations 👏 Thanks!

JD Santos 25 days ago

2
💻

BoltAI Mac v2

Planned

Custom providers should have custom icons

I’d like to set custom icons for my custom providers, so I can usefully enable icons in the sidebar. I frequently switch between different LAN-based “local” providers, but they are all the same “custom OpenAI-compatible” provider, so they all get an OpenAI icon. That doesn’t help me distinguish them from true external, paid, OpenAI.

Robert Meerman 27 days ago

1
💻

BoltAI Mac v2

Model Fusion mode

I would love to have model fusion mode. Basically what it does: Run multiple models side by side, run an analysis, and fuse into the best result. So, for example: I’m looking for the best AI tool for macOS... with XYZ features. Then it queries a few models (which you select), and as an output you get one combined prompt (that is combined by one model) Check this feature in action on openrouter (fusion) https://openrouter.ai/labs/fusion It works really well, and i’m starting to use it more and more! would be amazing to have it in BoltAI

David 28 days ago

1
💻

BoltAI Mac v2