Providers & Models

Cloud providers compared

Groq is fastest, OpenAI is the gold standard, Deepgram leads on streaming. All optional. Bring your own key.

Dimmy supports 8 cloud providers, plus a Custom endpoint for any OpenAI-compatible API. You bring the key, Dimmy talks to the API. We never proxy your traffic.

Provider	Type	Free tier	Best for
Groq	STT + LLM	Yes, rate-limited	Recommended. Sub-second on Whisper Large turbo, very fast Llama 3.3 70B.
OpenAI	STT + LLM	Pay-per-use	Highest accuracy on noisy audio (whisper-1). GPT-5.5 for LLM.
Deepgram	STT only	$200 free credits	Real-time streaming. Lowest end-to-end latency.
Gemini	STT + LLM	Yes	Long-context LLM (Gemini 3.1 Pro). Gemini 3.5 Flash for speed.
Anthropic	LLM only	Pay-per-use	Claude Opus 4.8 or Sonnet 4.6 for the LLM pass.
OpenRouter	LLM only	Yes (free models)	Try any open-weight LLM (Llama 3.3, DeepSeek, ...). One key, 100+ models.
Together	STT + LLM	Pay-per-use	Llama 3.3 70B Turbo. Parakeet TDT v3 for STT.
Fireworks	STT + LLM	Pay-per-use	Kimi K2 for LLM. Whisper v3 Turbo for STT.
Custom	STT + LLM	Yours	Any OpenAI-compatible URL. Self-hosted, alternate provider.

Adding a provider

Open Settings → Providers & keys and find the provider card. Add your API key in the key field. The key is stored encrypted on your device. See Add your API key. The On-device card on the same page lists your local models with a green check next to the ones already downloaded, see Choose a Whisper model.

Mixing local + cloud

You can run STT locally and LLM via cloud (or vice versa). For example: local Whisper for transcription (private), Claude Haiku for the LLM rewrite. Settings let you pick independently per pipeline stage.