Setup & Configuration
AnythingLLM is a full-stack app for running LLMs with RAG on your documents.
Supported LLM Providers
| Provider | Type |
|---|---|
| OpenAI | Cloud API |
| Anthropic | Cloud API |
| Ollama | Local |
| LM Studio | Local |
| LocalAI | Local |
| Azure OpenAI | Cloud API |
Embedding Engines
| Engine | Notes |
|---|---|
| Built-in | Default, no setup needed |
| OpenAI | text-embedding-ada-002 |
| Ollama | Local embeddings |
| LM Studio | Local embeddings |
Document Management
| Format | Support |
|---|---|
| Full text extraction | |
| DOCX | Full text extraction |
| TXT/MD | Direct ingestion |
| CSV | Tabular data parsing |
| Audio | Transcription via Whisper |
| Web URLs | Scrape and embed |
Workspace Settings
| Setting | Description |
|---|---|
| Chat mode | Query (RAG) or Chat (general) |
| Temperature | 0-1, controls creativity |
| Similarity threshold | Document match relevance |
| Max tokens | Response length limit |