We're excited to announce our collaboration with
@LangChainAI
!!!
The new Bearly Code Interpreter tool lets your ๐ฆ๐ agents safely execute code in a sandboxed environment. Check it out:
Just wrapped up a big month at !!
We had a few big product releases and dropped a bunch of new features - HyperChat, Encrypted Cloud Sync, our iOS App and more. Hereโs what's new:
๐จ New market map!
Our team
@a16z
consumer dove into AI x productivity tools. These products don't just help you organize tasks...they complete them for you.
What we're excited about ๐
Bearly AI enterprise plan now offers:
โข Single sign-on (SSO)
โข Privacy-first (zero-logging API queries)
โข Search and query large PDF collections
โข Multi-model access (Anthropic, OpenAI, Stability Diffusion, Kensho Scribe)
Fill this out for more:
Saturday was the best-selling day for . It happened after ChatGPT announced a $42 a month pro plan.
What does it mean?
I have no idea but my co-founder has been bugging me to tweet about it, so there you go.
Speech to text with transcript interactions now LIVE with
@bearly
- generate meeting minutes and immediately get the information each team member needs
- identify key takeaways across customer interviews
- ask specific questions of podcasts or YouTube videos
@Whisper
available
5/ โญ AI Frontend Startups
These are companies like
@jasper_ai_
and
@copy_ai
that are vertical specific โprompt engineering layersโ on top of the foundation model players.
These companies will use APIs from the foundation model players to stitch together end user solutions.
We're building the world's most private conversational AI. That means:
- Zero logging: we don't store or record your queries
- End-to-end encryption: you own the encryption keys. No one but you can access your interactions or docs
We're excited to showcase our youtube transcription feature. Paste a youtube link and start interacting with the content within Hyperchat. Absolute time saver!
@Anonyma_z
@MeHonestPerson
Sorry we missed this. Current limits are unlimited ChatGPT, 1000 image generations, 1M tokens GPT3.5 or 600K GPT-4 all reset monthly.
Limits are mostly to prevent automated use. In the ultra rare instances folks have hit it so far we've bumped their limit for free.
@Xanares_
@TrungTPhan
Deep Summaries and a few of our other prompts are multi-step GPT pipelines. They're chunked and processed and stitched. So no 4000 token limit.
๐ฎOur first big release was Bearly HyperChat.
HyperChat brings together all of our read, write, and analysis prompts into one central hub. Upload a doc โ> turn it into an agent โ> ask questions across any part.
Want to find key takeaways from analyst research reports?
@Aella_Girl
"The Drunken Jebus Party was started in 2006 by a group of college students who were looking for a way to let off some steam and have some fun..."
Nothing good starts this way.
๐ Weโve introduced Encrypted Cloud Sync in the spirit of continuing to make Bearly the most privacy-focused Chat experience in the world.
Generate your own key and no one but you can ever access the contents of your notebooks.
@fififofum
@TrungTPhan
Thanks!!! Goal is to get teams to encode their recurring tasks into great prompts/flows we help you write and share!
Same UI + team collaboration features and addons depending on the feature (e.g. Bearly Docs and sharing collections around teams when that launches next week).
@JustinBrady
@TrungTPhan
Weโre working with some great new technology to make this a reality. Speed is key factor to make this an enjoyable user experience.
In the meantime if your podcast is on YouTube, use the YouTube Summary.
@TheCryptoSlav1
@TrungTPhan
Thanks for the feedback Ivan! We're actively working on longform summary that will be available in the coming weeks. Podcasts, lengthy youtube videos and PDFs.
@PeterYeargin
@taps
@TrungTPhan
@pnegahdar
Peter! Invites are going out daily, and we'll get rid of that all together soon -- just making sure our systems can handle the load :)
Also an engineer is looking into the system events thing, we don't do anything interesting at OS level besides binding the hotkey.
@HeadandShouldR
Images are based on the StableDiffusion 1.5 (we can add 2.0 but no one really like 2 yet it seems). We will add Midjourney the second it becomes available!
@mattystevenson
Sure thing! Safari + FF are very soon, will be exactly like the chrome one.
Bearly routes to the best model available (currently latest gpt + other domain specific models). Weโll add support for the Google et al as the come
V2 will support DallE and Stablediffusuon as well.
@mattystevenson
No way!
Weโre just focused on making Bearly better and bringing even more value ๐คฉ
Btw we have some cool drops this week, weโre looking forward to hearing your thoughts!
@jcubic
@wahVinci
Side note for the privacy thing: we don't log or store any of your queries and handle most things client sided (though OpenAI has a 30 day retention policy but all our users request go through the our api keys so theres some anonymization there as well).
@PradSagar
Hey!!
While accept Indian credit cards, we're thinking about how to approach pricing in different regions (currently every country is $20 USD/month for pro). If you have any suggestions please DM us!
@kiranadimatyam
@LangChainAI
Bearly's has improved quite (we do weekly releases) a bit and we still have a long way to go! Hope you give us another try and let us know what you think!
Please DM me or
@pnegahdar
if you have any specific requests or issues.
@mattystevenson
Our roadmap is truly wild! Extension V2 was submitted for review, should be out tomorrow.
App v2, with a ton of new things, is coming out this week!
Our main goal: keep pushing you the worlds best ML via an interface that actually integrates in your workflow.