Providers¶
QUASAR supports multiple AI providers for flexibility and reliability.
Supported Providers¶
| Provider | Type | Best For | Free Tier |
|---|---|---|---|
| Cerebras | Cloud | Tool calling, reliability | ✅ |
| Groq | Cloud | Speed, throughput | ✅ |
| Ollama | Local | Privacy, offline use | ✅ |
Provider Priority¶
QUASAR automatically selects providers in this order:
1. Cerebras (best for tool calling)
2. Ollama (local, reliable)
3. Groq (fast, but some tool issues)
If a provider fails or hits rate limits, QUASAR automatically falls back to the next.
Override Provider¶
Use --model to specify a provider:
quasar --model cerebras/zai-glm-4.7 "your request"
quasar --model groq/openai/gpt-oss-120b "your request"
quasar --model ollama/qwen2.5-coder:7b "your request"
Format: provider/model-name