@Tech2Wild Vision improves things a lot, for me, so I'd take a speed hit to run a vision model.. personally.
5.3 Flash is such a huge leap above DS4F for me.. idk. I hope people can figure out ways to speed it up 🤞
@merge_api Pretty rough experience so far with whoever Particle AI is. Had a session respond in both Chinese & Indonesian to english prompts.. randomly, mid chat. lol.
@Satoshi87164293@MiaAI_lab@Zai_org@opencode They have different limits per model, until they get a better inference deal you only get $15 on 5.3/5.3-Flash
https://t.co/j75wCORQwZ
@NoemiTitarenco Scroll down to the next table in usage limits, shows you the exact dollar value usage you get per model. Not all are $60 unfortunately.
@NoemiTitarenco Scroll down to the next table in usage limits, shows you the exact dollar value usage you get per model. Not all are $60 unfortunately.
@pmwoz@jun_song@theo It's the retail pricing for flash now. Weights aren't published yet so there's no one else hosting quantized versions for dirt cheap.
@maria_rcks 3.7 has been decent at some front end stuff I threw at it the last week and it's fast.. 🤷♂️
I treat gemini as free tokens with my Google storage upgrade.
@MariaMartinsPT@0xSero Did you login with a free https://t.co/Dco3Qieh5c account? I got it last weekend and just checked and have it again this weekend.