GPT-6.1 Sol (and Sol Pro)

GPT-6.1 Sol (openai/gpt-6.1-sol) is the update to OpenAI’s balanced tier: same price and contract as GPT-6 Sol, with refined reasoning. The openai/gpt-6.1-sol-pro variant keeps the same capabilities and pricing for higher workloads.

Parameters and capabilities

Parameter	Value
model	openai/gpt-6.1-sol, openai/gpt-6.1-sol-pro
Images on input	yes (image_url, data URLs)
Files on input	yes (PDF/DOCX/TXT etc., up to 5)
Reasoning	reasoning_effort: low, medium, high, xhigh, max; cannot be disabled
Native web search	yes — web_search: "native"
Cache read	$0.125 per 1M tokens
Cache write	$3.125 per 1M tokens
seed	yes — best-effort sampling reproducibility

The difference from GPT-6 Sol: on the 6.1 generation reasoning cannot be disabled — a request with reasoning_effort: "off" is rejected. GPT-6 Sol and Luna did support off.

Example

Python

from openai import OpenAI

client = OpenAI(base_url="https://api.mixen.ai/v1", api_key="mxn-...")

resp = client.chat.completions.create(
    model="openai/gpt-6.1-sol",
    messages=[{"role": "user", "content": "Design a multi-tenant SaaS database schema and justify the choices"}],
    reasoning_effort="high",
)
print(resp.choices[0].message.content)

cURL

curl https://api.mixen.ai/v1/chat/completions \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-6.1-sol",
    "messages": [{"role": "user", "content": "Design a multi-tenant SaaS database schema and justify the choices"}],
    "reasoning_effort": "high"
  }'

Web search — via "web_search": "native"; sources arrive in message.annotations.

Pricing

Per 1M tokens: input — $2.5, output — $12.5 — same price as GPT-6 Sol. Cache: read — $0.125, write — $3.125. Showcase prices follow the live exchange rate — take fresh values from the catalog.

Strengths and limitations

• A drop-in replacement for GPT-6 Sol: change model and nothing else; the contract and price stay.
• Reasoning is always on: thinking tokens are billed as output tokens, budget for them in max_tokens.
• Native web search and file input — like the whole GPT-6 family.
• The Responses protocol (POST /v1/responses) works with the same key — for Codex CLI and the OpenAI Agents SDK.

The GPT-6 family — model card. All models — in the catalog. The general workflow — chat guide.