GPT-5.5 pro
OpenAI's strongest reasoning model — extended thinking for the hardest math, science, and multi-step problems. Call it through the OpenAI-compatible POST /v1/chat/completions — swap the base URL on your existing OpenAI SDK and pay per call. Billing is a pre-charged ceiling: (estimated input tokens + your max_tokens) at the per-1M rates, so set max_tokens to bound spend; the response returns real usage.
Quick start
# Inspect the price — a plain request returns the 402 challenge:
curl -i https://api.glianalabs.com/v1/chat/completions \
-H "content-type: application/json" \
-d '{
"model": "openai/gpt-5.5-pro",
"messages": [{"role": "user", "content": "<your message>"}]
}'
# Pay + run in one step with the mppx CLI (create a wallet: npx mppx account create):
npx mppx https://api.glianalabs.com/v1/chat/completions \
-J '{"model": "openai/gpt-5.5-pro", "messages": [{"role": "user", "content": "<your message>"}]}'Parameters
OpenAI-style chat messages: [{ role: "user" | "system" | "assistant" | "tool", content }]. Vision: content may be a parts array with { type: "image_url", image_url: { url } } (public URL — upload via POST /v1/media).
Output cap — bounds the pre-charged ceiling. Default 1024.
Sampling temperature. Passed through where the provider supports it.
Nucleus sampling. Passed through where the provider supports it.
true → Server-Sent Events (OpenAI chunk format) on /v1/chat/completions.
OpenAI function tools: [{ type: "function", function: { name, description, parameters } }]. Response returns tool_calls (finish_reason "tool_calls"); send results back as { role: "tool", tool_call_id, content }.
"auto" (default) | "none" | "required" | { type: "function", function: { name } }.