⚠️ Bot & API is currently down for maintenance. We are in the process of switching to a new domain & adding new models (Fable). ETA: 1-2 hours.

📢 Bot & API is fully functional again after a short period of downtime

✅ Added over 85 models (Claude Fable 5, GPT 5.5, DeepSeek v3.2, GLM-5, Kimi K2.5, etc) at $0.5/1M tokens

ℹ️ Switched domains, regenerate your config in "How to setup" to get the new domain (old domain still works for now)

We've discovered an issue with our token tracking since the update on July 17th and fixed it. Users who were overcharged for cached tokens have been fully compensated. The API should be up to 90% cheaper now, give it a try!

Opus 5 has been added! Same price as previous Opus ($0.5/1M)

🔥 Prices have been reduced by 10%, all models are now just $0.45/1M!

🔥 GPT 5.6 Sol, Terra, and Luna have been added at the same price of $0.45/1M, the cheapest rate on the market. Generate an OpenCode config in the bot to use.

Added deepseek-v4-flash-0731 and kimi-k3. Chinese models may be unstable atm.

Claude models are temporarily down due to issues on our upstream's side and will be fixed shortly.

Claude models are back up and functional.

Huge banwave on Claude, most providers were hit by this. Use GPT & Chinese models in the meantime.

Claude models were fully restored as of 45 minutes ago.

Claude models are extremely unstable due to issues on all upstreams, every provider around our price point is having issues atm. Please use GPT 5.6 Sol, GLM-5.2 etc in the meantime. They are stable.

Claude has been back up and working very stable for the past 24 hours. We also pushed an update to automatically switch upstream providers if the current one dies or becomes unstable, this should prevent further outages.

Added Grok 4.6, Grok 4.5, DeepSeek v4 Pro (08/13, newest release). OpenCode configs were also updated to add customizable thinking, regenerate yours if you use it.