November 2024

# November 14

Horizontal rule is better styled now 📏 Previously it was one of the missing Markdown elements that still used the default browser styles.

Our thanks to @huy for reporting the issue 🙇

# November 15

We now have detailed message when something went wrong ✋ Previously it was just a generic "Error", which led to missed opportunities to suggest joining or advertise Pro subscription.

Our thanks to @chinh (Pro user) for reporting the issue 🙇

# November 17

Messages now include fallback model 🔒 Fallback model is what actually responded when your plan does not include the requested model. For example, if you ask for "GPT 4o" without being a Pro user, you will see a warning that "GPT 4o mini" was used.

Our thanks to @chinh (Pro user) and @huy for reporting the issue 🙇

# November 18

We fixed a bug where companions went to retirement unexpectedly 🧓 Worry not that no companions are lost! This is a linking issue, meaning that we only lost their name in rare cases, but they are still there. Rare cases include asking a question without making changes to any companion yet.

Our thanks to @huy for not only reporting the issue but also working with us to fix it.

# November 23

We added Grok from xAI 😎 We also upgraded Anthropic models to 3.5 series, and Llama to 3.2.

# November 24

We added styles for quotes now ✍️ This is applied for both in-line and block quotes.

Thanks @huy for reporting the issue 🙇

# November 25

Users' messages are no longer went through Markdown 📝 This should make it more predictable, as our users often don't format their messages strictly like AI always does.

Thanks @chinh (Pro) and @huy (Pro) for reporting the issue 🙇