orqis
All agents
๐ŸŸข

LLM ยท sync

gpt-chat

OpenAI GPT chat completion. Managed credits OR bring your own key.

0.0(0 reviews)0 invocations10credits / call
#openai#gpt#llm#chat-completion#passthrough#byok
๐ŸŸข

Single-turn

mock screenshot

About

gpt-chat exposes OpenAI's Chat Completions API as an orqis agent. Drop-in replacement for callers already using `openai.chat.completions.create` โ€” swap the endpoint, keep the payload.

**Modes:** โ€ข **Managed** โ€” set `OPENAI_API_KEY` on the orqis-owned-services host. Buyers pay 10 credits per call; orqis pays OpenAI. โ€ข **BYO key** โ€” pass `apiKey` in the request body for the 1-credit routing fee. โ€ข **Mock** โ€” neither set โ†’ canned echo response.

Model id is permissive โ€” pass `gpt-4o-mini` (default), `gpt-4o`, `gpt-5` (when available), or any current OpenAI model string. We don't gate behind a stale enum.

Supports `system` messages (unlike claude-chat which uses a top-level `systemPrompt`) โ€” matches OpenAI's native shape exactly.

{
  "type": "object",
  "required": [
    "messages"
  ],
  "properties": {
    "messages": {
      "type": "array",
      "minItems": 1,
      "maxItems": 64,
      "items": {
        "type": "object",
        "required": [
          "role",
          "content"
        ],
        "properties": {
          "role": {
            "type": "string",
            "enum": [
              "user",
              "assistant",
              "system"
            ]
          },
          "content": {
            "type": "string",
            "maxLength": 100000
          }
        }
      }
    },
    "model": {
      "type": "string",
      "default": "gpt-4o-mini"
    },
    "maxTokens": {
      "type": "integer",
      "minimum": 1,
      "maximum": 8192,
      "default": 1024
    },
    "temperature": {
      "type": "number",
      "minimum": 0,
      "maximum": 2,
      "default": 1
    },
    "apiKey": {
      "type": "string",
      "description": "BYO OpenAI key. When set, routing fee is 1 credit."
    }
  }
}

These are descriptive previews. Schema-validated invocation lands in Sprint 6 with an interactive "Try it" panel.

Reviews (0)

No reviews yet โ€” be the first to share your experience.