ganso

[ganso @gansotragico]

@Samking207 @SaeedSardarKhan @Zai_org To be fair I had a similar experience at some point. Customer service is lacking a bit.

[ganso @gansotragico]

@thalynita Encima de que te paso las originales

[ganso @gansotragico]

@thalynita Y esta también la meten por el aniversario.

[ganso @gansotragico]

@thalynita Es fake, esta es la original.

[ganso @gansotragico]

@thalynita Porque las bolsas pesan más que tú vale si jejeje

[ganso @gansotragico]

@opencode I use Kimi K2.6 not only through OpenCode, but also through Kimi Code with their coding plan, and it is also completely unresponsive at the moment.

[ganso @gansotragico]

@thegenioo @OpenAIDevs @thsottiaux While GPT-5.5 added increased quality it should not be used for most simple / small tasks.

Even though it’s more efficient with tokens that is something you’d only be able to see with longer horizon tasks from what I’ve been testing so far

[ganso @gansotragico]

@LucasVHoutven @opencode You are using it through Kimi and not OpenCode Go, that’s why

[ganso @gansotragico]

@brendonovich @notshivmsingh WSL integration would be really cool. As of right now, we need to use the `opencode web` command from WSL terminal which is fine but not exactly the same.

[ganso @gansotragico]

@boy_key_666 @opencode Maybe they are serving both Qwen and MiniMax through Alibaba so they share infrastructure and hence the lower quota :(

[ganso @gansotragico]

@Zai_org @NVIDIAGTC I use GLM-5-Turbo via OpenCode to auto-generate Architecture Decision Records from PR diffs with built-in compliance validation. Catches security issues at commit time instead of audit season.

[ganso @gansotragico]

@Zai_org GLM Coding Plan is straight up unusable right now (especially GLM-5). Time to first token is way too high and inference speed makes tool calling through CLI options like OpenCode / ClaudeCode painfully slow.

Is there any plan to improve this in the future?

[ganso @gansotragico]

@MiniMax_AI I’ve been using the latest models to orchestrate sub-agent workflows in OpenCode, and so far MiniMax has been the best experience I’ve had.

Tool calling is seamless, reasoning is strong, and throughput is excellent in both MiniMax M2.5 and MiniMax M2.5 HighSpeed.

[ganso @gansotragico]

@thalynita Sigue siendo gracioso

[ganso @gansotragico]

@RL_Status @the_mical Also happening on my end. I don't know what's going on but it's happening quite frequently lately. Not just me. Same applies to a couple friends.

[ganso @gansotragico]

@sama What about emojis? I cannot seem to get rid of those no matter what. It always uses emojis when calling the web search tool.

[ganso @gansotragico]

@Trae_ai @RuthesDev The post request being sent when joining the waitlist throws an exception and never goes through, so it’s basically not possible to join :(

[ganso @gansotragico]

@thalynita wtf jajajaja

[ganso @gansotragico]

@thalynita Que chulo ese bug. Es skater jejejej

[ganso @gansotragico]

@thalynita Si que hizo cosas jejejeje