Meet Limbani

Fail Faster , Learn Quicker

LinkedIn:

[Meet Limbani @winds_ai]

@poteto Crazzyy news and in crazyy timeframe

[Meet Limbani @winds_ai]

Hi @cartesia team, how do i avail the 100$ free credits for me?

[Meet Limbani @winds_ai]

I would like to get hands on that free 100$ credit

[Karan Goel @krandiash]

We released Sonic-3.5 and Ink-2, the #1 streaming models for text to speech and speech to text you can use in your voice agents today.

New architectures enable new frontiers for speed and quality.

We're now the only provider to have #1 models for both speaking and listening.

[Meet Limbani @winds_ai]

some are just doing it for engangement farming and some are just influned by their favorite creater or dev.

Honetly both usecases are more clearer to people now because earlier we were trying to use both for same problem that needs either one only.

and one random take that all docs targeted to agents should just be cli now instead of webapges.

in cli there is not scope of hidden html or css fonts and they already know cli commands and have access to clis mostly

[Meet Limbani @winds_ai]

@petradonka Yuppp sooo true, this is the same reason that we haven't fully utilised the capabilities of latest sota yet and before we would - new sota will be released

[Meet Limbani @winds_ai]

@weswinder @thsottiaux @MehakdeepK81 In go, so I assume that for free it will be less then go

[Meet Limbani @winds_ai]

@weswinder @thsottiaux @MehakdeepK81 I had a spare go account so I just tried Yesterday in pi with minimal reasoning and gpt 5.5

All in all we get 3-4$ of inference in go per month

[Meet Limbani @winds_ai]

@reach_vb Come to Ahmedabad vb.... Some of are waiting

[Meet Limbani @winds_ai]

@og_brethren There is only monthly resets in go

[Meet Limbani @winds_ai]

you get about 3$ of codex inference in openai's go plan at this moment

[Meet Limbani @winds_ai]

@jerryjliu0 Yuppp more then next generation of models we need ecosystem around those models more

[Meet Limbani @winds_ai]

@thsottiaux @MehakdeepK81 Please try using the free one if you haven't recently tibo.

It doesn't have enough inference to o any real work in codex

[Meet Limbani @winds_ai]

@ivanfioravanti @OpenRouter Yeahhhh this is what many people were trying but only a company like openrouter can verify this idea with benchmarks and serve with multiple models easily

[Meet Limbani @winds_ai]

@argofowl for pro plans also?

[Meet Limbani @winds_ai]

@Dimillian Can you share the screenshot of that

Totally not to check whether you're using gpt 6 or not

[Meet Limbani @winds_ai]

@argofowl 70% chance based on intuition

[Meet Limbani @winds_ai]

@jxnlco It regularly misses that there is a header on top and many times the updates and tsv payloads are just 1 row or column off and it messes up the data

[Meet Limbani @winds_ai]

@thsottiaux @daniel_mac8 yayyy or nayyy?
https://t.co/kOfPYbPcXo

[Meet Limbani @winds_ai]

I want to suggest a update for compaction.

can you guys stop the agent after calling compaction if credits are exhausted.

suppose i had a long running task that was stopped because 5h limit exhausted so now when i come back when session refreshed then it takes longer time because cache is not present and also consume usage more even if that cache is not going to be useful from then on

[Meet Limbani @winds_ai]

@jxnlco Broooooo
https://t.co/kOfPYbPcXo

[Meet Limbani @winds_ai]

I want to suggest a update for compaction.

can you guys stop the agent after calling compaction if credits are exhausted.

suppose i had a long running task that was stopped because 5h limit exhausted so now when i come back when session refreshed then it takes longer time because cache is not present and also consume usage more even if that cache is not going to be useful from then on

[Meet Limbani @winds_ai]

@Dimillian might also save some noticable amount of compute being wasted right now because of this and can actually be used

[Meet Limbani @winds_ai]

I want to suggest a update for compaction.

can you guys stop the agent after calling compaction if credits are exhausted.

suppose i had a long running task that was stopped because 5h limit exhausted so now when i come back when session refreshed then it takes longer time because cache is not present and also consume usage more even if that cache is not going to be useful from then on