OpenAI

Use GPT-4o, GPT-4o mini, o1, o3-mini, and any future OpenAI model for scans and fixes. One key, streaming support, tool calling.

Preset setup — 60 seconds

Settings → Models → Add custom provider
Select preset: OpenAI
Paste your sk-... from platform.openai.com/api-keys
Choose model: gpt-4o (recommended) or gpt-4o-mini for cost.
Test connection → Save → Set as Preferred

Auto-filled values for OpenAI preset:

Base URL: https://api.openai.com/v1

Auth: Bearer $OPENAI_API_KEY

Streaming: true, Tools: true

Manual config

If you don’t use preset, fill these:

Field	Value
provider	openai
baseURL	https://api.openai.com/v1
apiKey	sk-proj-...
model	gpt-4o
extra	{ stream: true, temperature: 0.2 }

Model notes

gpt-4o$$

Best all-around for fixes. High accuracy patching, understands large diffs.

gpt-4o-mini$

Budget scans. Good for low/info findings, faster.

o1 / o1-mini$$$

Deep reasoning for complex vulnerabilities (SSRF, auth bypass). Slower.

o3-mini$$

New reasoning model, better than o1-mini for code.

Direct API test

Use this to verify your key before adding to Rixel:

curl https://api.openai.com/v1/chat/completions \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{ "role": "user", "content": "Say hello" }]
  }'

Troubleshooting

401 Unauthorized: Key invalid or project not enabled for model. Check organization → enable model.

429 rate limit: Add payment method or raise tier at platform.openai.com/account/limits.

Model not found: Your account may not have access to o1/o3. Use gpt-4o first.

Next: Anthropic →