> ## Documentation Index
> Fetch the complete documentation index at: https://docs.scitix.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# DB-GPT

> Learn how to configure DB-GPT with ScitiX Model Inference for chat-to-data, agents, and RAG data applications.

[DB-GPT](https://github.com/eosphoros-ai/DB-GPT) is an open-source AI-native data app framework (chat-to-data, agents, RAG). Its `proxy/openai` model provider targets any OpenAI-compatible endpoint, so it runs on the ScitiX Model Inference API.

Default model here is `glm-5.2`; see the [Models](https://console.scitix.ai/model-inference/models) page for the full list and pricing.

Deploy DB-GPT per the [official guide](https://docs.dbgpt.site) — the lightweight `dbgpt-openai` image is the proxy-only build (no local GPU model) and is the right base for ScitiX.

## Configure

Models in DB-GPT are defined server-side in a TOML config, not in the UI. Use the `proxy/openai` provider and the shipped `configs/dbgpt-proxy-openai.toml` config. This is the path for OpenAI-compatible endpoints like ScitiX, and the proxy-only `dbgpt-openai` build needs no local GPU model.

The config already reads every value from the environment, so point it at ScitiX by setting these variables. Start the server with `--config configs/dbgpt-proxy-openai.toml`.

```toml theme={null}
# configs/dbgpt-proxy-openai.toml (excerpt — values come from env)
[[models.llms]]
name = "${env:LLM_MODEL_NAME:-gpt-4o}"
provider = "${env:LLM_MODEL_PROVIDER:-proxy/openai}"
api_base = "${env:OPENAI_API_BASE:-https://api.scitix.ai/model-api/v1}"
api_key = "${env:OPENAI_API_KEY}"

[[models.embeddings]]
name = "${env:EMBEDDING_MODEL_NAME:-Qwen/Qwen3-Embedding-8B}"
provider = "${env:EMBEDDING_MODEL_PROVIDER:-proxy/openai}"
api_url = "${env:EMBEDDING_MODEL_API_URL:-https://api.scitix.ai/model-api/v1/embeddings}"
api_key = "${env:OPENAI_API_KEY}"
```

Create an API key on the ScitiX [API Keys](https://console.scitix.ai/model-inference/api_keys) page, then set:

```bash theme={null}
LLM_MODEL_NAME=glm-5.2
LLM_MODEL_PROVIDER=proxy/openai
OPENAI_API_BASE=https://api.scitix.ai/model-api/v1
OPENAI_API_KEY=<Your API Key>

# Embeddings (RAG / Knowledge Base) — point at a ScitiX embedding model
EMBEDDING_MODEL_NAME=Qwen/Qwen3-Embedding-8B
EMBEDDING_MODEL_PROVIDER=proxy/openai
EMBEDDING_MODEL_API_URL=https://api.scitix.ai/model-api/v1/embeddings
```

| Field                     | Value                                                                                                    |
| ------------------------- | -------------------------------------------------------------------------------------------------------- |
| `LLM_MODEL_NAME`          | ScitiX model ID, e.g. `glm-5.2`. Pick a model with native tool calling for DB-GPT agents and data tools. |
| `LLM_MODEL_PROVIDER`      | `proxy/openai`                                                                                           |
| `OPENAI_API_BASE`         | ScitiX OpenAI-compatible endpoint: `https://api.scitix.ai/model-api/v1`                                  |
| `OPENAI_API_KEY`          | Your ScitiX API key                                                                                      |
| `EMBEDDING_MODEL_API_URL` | ScitiX embeddings endpoint: `https://api.scitix.ai/model-api/v1/embeddings`                              |

## Verify

Start the web server and open the UI (default `http://localhost:5670`). In a chat, the model selector shows **glm-5.2** — send a message and a normal reply confirms the connection.

<img src="https://mintcdn.com/scitix-adfc65c2/AblxxMM8pLmVYmdt/model-inference/media/db-gpt-chat-reply.png?fit=max&auto=format&n=AblxxMM8pLmVYmdt&q=85&s=57e05229014fd5c588ca17c6f0f3f98f" alt="DB-GPT chat on glm-5.2 served through ScitiX" width="2988" height="1436" data-path="model-inference/media/db-gpt-chat-reply.png" />

The reply comes back as the "DB-GPT assistant" persona. That is DB-GPT's own agent framing, not a routing issue; the `Model: glm-5.2` label under the message confirms the ScitiX model in use.

Verified on DB-GPT v0.8.1.