AI Providers
Youkore needs an LLM model to think — its brain. To use one you need an API key from a provider, or a local runtime (LM Studio, Ollama) that exposes a compatible endpoint.
The setup wizard walks you through the exact steps for the provider you pick. This section is the same content, kept online so the instructions stay up to date as provider consoles change — no app update needed.
Cloud providers (API key required)
| Provider | Free tier | Notes |
|---|---|---|
| Anthropic (Claude) | No | Prepaid credits, $5 enough to start |
| OpenAI | $5 credit | Phone verification at signup |
| Google (Gemini) | Yes | OpenAI-compatible endpoint |
| Groq | Yes | Very fast, no card required |
| xAI (Grok) | No | API activation may take a few hours |
| DeepInfra | No | Pay-per-use, many open-source models |
| Moonshot | No | Kimi K2.5, 256K context |
| Together AI | No | Many open-source models |
| Mistral AI | Yes (Experiment plan) | Billing must still be enabled |
| Cerebras | Yes | 1M tokens/day, fastest inference |
| Fireworks AI | Limited | 10 RPM without a card |
| OpenRouter | No | Aggregator: hundreds of models, one key |
Local runtimes (no API key)
| Runtime | Notes |
|---|---|
| LM Studio | Desktop app, free, GPU recommended |
| Ollama | CLI, free, NVIDIA/AMD/Apple Silicon |