@kimmonismus I just see my weekly allowance is draining 10 times faster than usual on a 200/month plan!
My weekly quota reset 4 hours ago and I am now at 62% while normally I would be at ~97% given my modest usage in the last few hours.
@thsottiaux
@frydwia ok but still the whole product is literally based on a "chat"...I mean sending messages to viktor in slack/teams is still texting...there is simply no separate UI, but it's still the core
The speed of Gemini 3.7 is one of the most damning things about it.
Its t/s is so much higher than the competition, and yet it gets speedmogged by GPT 5.6 Sol when it comes to actually finishing the task you gave it.
This doesn't even take into account fast mode.
@reddygtvs@ToonamiAfter@_philschmid Then 3.7 is even less token efficient…so they are literally compensating for extreme verbosity so the actual user perceived speed is probably quite bad or average:
Gemini-3.7-Flash is way less token efficient than the already embaressingly inefficient 3.6-Flash but to compensate Google made it like 3x cheaper per token evening out to still being a lot cheaper
But this likely means Google coudlnt figure out how to make it actually smarter so they just had it burn more tokens and they cut down on their massively greedy original profit margins a little to be less embaressing
@reddygtvs@ToonamiAfter@_philschmid Yes, it’s not the only metric but it’s one of them…you’d expect sonnet 5 to be quite slow model…it is, but definitely not as much across use cases, especially shorter ones…
After trying every single gemini models up to now, I will not even test 3.7. I decided to stop wasting time on this.
I will just read X reviews.
But you know it's another useless pumped-up and model with misleading stats when as usual only google employees brag about it.
Introducing Gemini 3.7 Flash : )
- it is fast!
- 50% lower price than 3.6 flash (through end of year)
- strong intelligence increase in only ~3 weeks (thanks to some awesome algorithmic improvements)
- available in the API, AI Studio, Antigravity, and more!
@ToonamiAfter@_philschmid they live off such misleading stats...tokens per second means nothing if initilzation time is way larger than other anthropic/openai...
Hey @AskWorkspace , Navor Sàrl cannot redeem its Google for Startups 12-month Workspace Business Plus offer. Your provided confirmed promo code is rejected, and our new admin email is redirected to checkout with no support access.
Please help us escalate this!!!
@Nick_zv_ Interesting pulling the site map of competitors.
But, especially in presence of emerging trends, isn’t the advice to target new keywords derived by new habits that competitors don’t focus much on? Do we need semrush/agree for that or there is a free way?