Otras muchas veces el modelo piensa, pero nisiquiera contesta.
Desde la app del móvil, seleccionar el modelo es horrible, estaría bien tener un buscador de modelos en el que poder escribir el nombre.
Todas mis pruebas están hechas con OpenCode, en terminal funciona bien.
He estado resistiéndome de usar t3code bastante tiempo y por fin me he animado, pero la verdad es que estoy bastante decepcionado con lo mal que va.
Muchas veces se corta la contestación del modelo, dejando la conversación a mitad.
@maria_rcks
being pro-opensource doesn't mean you have to be anti-proprietary
it's fine for proprietary options to exist, they usually invent the category and make it possible for the opensource option to emerge
the only thing that matters is the opensource option is allowed to exist
@thdxr I'd been working on it for years, but AI has sped up the process. It's a Discord bot for role-playing, similar to Dungeons and Dragons but adapted for Discord.
Here's an example of the character sheet and 2D combat
@XMihura Estaria divertido hacer el mismo experimento pero como fue realmente. Hacer varios equipos cada uno con un modelo, 5.6 Sol, Fable 5, Muse Spark 1.1, Grok 4.5 y GLM 5.2 intentando mejorar las soluciones del otro
the $200/mo Codex plan used to be a steal
I never hit the limits, even running multiple /goal loops with GPT 5.5 xhigh
I could literally have 10+ agents running in parallel and would NOT hit the limit. ever.
but now, with GPT 5.6 Sol, i hit the limits instantly
even while running just 1-2 agents
i'm not happy with this at all...
@thsottiaux Is it normal for the ChatGPT app to be consuming 4GB of RAM and to have burned through my entire 5-hour Plus plan quota on a single prompt, using 5.6 terra just for a trivial UI change?
the funniest outcome would be:
Zai, DeepSeek, or Qwen releasing an opus 4.8 / gpt-5.6-level model and open-sourcing it before either of them even releases theirs publicly
that would be the perfect ending to this whole "holding back frontier models" circus