DaoXEBecause Roo Code accepts any OpenAI-compatible endpoint, moving it to DaoXE is a base-URL-and-key change. Roo's twist is per-mode model profiles — so you can run a frontier model in Architect mode and a cheaper one in Code mode, all on one key. See the cheapest APIs guide for where savings really come from.
Why route Roo Code through DaoXE#
- Per-mode models. Assign different DaoXE models to Code, Ask and Architect modes.
- One key, many vendors. No separate accounts for GPT, Claude, Gemini, DeepSeek or Kimi.
- Drop-in. If you already use Cline, the same settings apply — Roo is a fork.
Set it up in Roo Code#
- Open Roo settings. Click the gear in the Roo Code panel, then under Providers choose OpenAI Compatible.
- Set Base URL & key. Base URL
https://daoxe.com/v1(end at/v1, no/chat/completions), then paste your DaoXE key. - Pick the Model ID. Enter an exact ID from
GET /v1/models; Roo can populate the list from your key. - Assign per mode (optional). Give Code, Ask and Architect modes different models to balance cost and quality.
Caveats worth knowing
The Base URL must not include /chat/completions — Roo appends the path. Agent actions need native tool calling; pick a tool-capable model or edits fail. Model IDs are account-scoped (read GET /v1/models). Keep separate provider profiles if you switch back and forth. More clients: client setup notes.
Verify you actually get the model#
Prove the endpoint works before blaming the client — if this fails, no setting will fix it:
export DAOXE_API_KEY="your_api_key"
# List the exact model IDs your account can call
curl --fail-with-body --show-error --silent \
https://daoxe.com/v1/models \
-H "Authorization: Bearer ${DAOXE_API_KEY}"Confirm the endpoint responds, then compare a hard prompt against the official API at temperature 0 to rule out a downgrade:
curl --fail-with-body --show-error --silent \
https://daoxe.com/v1/chat/completions \
-H "Authorization: Bearer ${DAOXE_API_KEY}" \
-H "Content-Type: application/json" \
-d '{
"model": "YOUR_EXACT_MODEL_ID",
"max_tokens": 64,
"messages": [{"role": "user", "content": "Say hello in one sentence."}]
}'Verify us — don't trust us
Point the open benchmark at DaoXE and at the official API and compare at temperature 0. Then learn to detect model swapping so a cheaper endpoint can't quietly swap you to a smaller model.
Frequently asked questions#
Is Roo Code setup the same as Cline?
Largely yes — Roo is a Cline fork. Choose OpenAI Compatible, set https://daoxe.com/v1, paste your key, and pick an exact model ID.
Can each mode use a different model?
Yes — Roo supports per-mode model profiles, so Architect can use a frontier model and Code a cheaper one.
Tool calls fail — what's wrong?
Roo needs native tool calling. Switch to a tool-capable model ID from your account.
Where do I find model IDs?
Call GET /v1/models with your key; copy exact IDs (they're account-scoped).
Try DaoXE — and benchmark it yourself
One key for GPT, Claude, Gemini, DeepSeek and more. Point the open benchmark at us and compare — don't take our word for it.