Groq gets a current model list that keeps itself up to date (#755).
- Models: without a refresh, Groq offers GPT-OSS 120B, GPT-OSS 20B and Qwen 3.8 27B. Llama 3.3 70B, Llama 3.1 8B and Kimi K2 are gone, because Groq retired them for free and developer accounts.
- Refresh models: the Groq connection settings have a Refresh models button again. The list also reloads in the background once a day, so a model Groq retires drops out on its own.
- Default: GPT-OSS 120B is the default. If your saved model is no longer offered, Groq uses GPT-OSS 120B instead of failing with an error, and your choice applies again once Groq lists the model again.
Thanks to @7Gamil for the model update.
Works with TypeWhisper for Windows 1.1.0 and later.
Source commit: d7116df