GPT-6 Sol API: Pricing, the 272K Cliff and Setup

GPT-6 Sol is OpenAI's coding and agent model in the GPT-6 family, live on RouterPlex as gpt-6-sol. GPT-6 Sol API pricing is $2 per 1M input tokens and $10 per 1M output tokens at or below 272,000 input tokens. Cross that line and the whole request bills $4 input and $15 output per 1M.

That is one fifth of GPT-6 Astra ($10 / $50) and twenty times GPT-6 Luna ($0.10 / $0.50) on the input side. OpenAI shipped both Sol and Luna on 23 September 2026. Astra stays the flagship.

GPT-6 Sol API pricing #

USD per 1M tokens. The first two rows are what RouterPlex bills. The cache row is the catalog cache-read rate, which is separate from OpenAI's published cached-input line.

OpenAI's model card, same day, lists a different cache sheet for the same sticker:

RouterPlex bills cache reads at $0.16 per 1M ($0.32 above 272K). That is the number on the live catalog. It is lower than OpenAI's published $0.20 cached-input rate. RouterPlex does not publish a separate cache-write price, a Batch price, or a Flex price. OpenAI's card prices Batch and Flex at 50% of Standard, Fast at 2×, and regional processing at a 10% uplift where that processing is available. This key bills the standard list rate in the table above.

A 40,000-in / 2,000-out turn with no cache costs $0.10. The same turn is $0.005 on gpt-6-luna, $0.50 on gpt-6-astra, and $0.26 on gpt-5.6-sol at the catalog rate ($5 / $30).

A $5 prepaid balance covers 2.5 million Sol input tokens, or 500,000 Sol output tokens, if a workload used only one side. Real requests mix both. Reasoning tokens bill as output.

The 272K cliff applies to the whole request #

Once input tokens go past 272,000, the higher rate applies to every token in that request. There is no blended tail.

271,000-token prompt on gpt-6-sol  →  271,000 × $2.00/1M  = $0.542
273,000-token prompt on gpt-6-sol  →  273,000 × $4.00/1M  = $1.092

Two thousand extra input tokens about double the input bill, and output on that same request moves from $10 to $15 per 1M. If a coding agent is anywhere near 272K, trimming the prompt below the threshold is worth more than a small cache tweak. The same rule is on Astra ($20 / $75) and Luna ($0.20 / $0.75).

GPT-6 Sol vs Luna, Astra, and GPT-5.6 Sol #

List prices on RouterPlex, USD per 1M tokens, and the cost of a 40,000-token prompt returning 2,000 tokens with no cache. The GPT-5.6 Sol row is the catalog rate, which is above OpenAI's $4 / $20 promotional list.

OpenAI's 23 September 2026 table shows GPT-6 Sol at $2 / $10 against a GPT-5.6 Sol promotional price of $4 / $20, and labels that a 50% reduction. RouterPlex bills the new Sol list. It still bills GPT-5.6 Sol at $5 / $30. If you are costing a migration off the old Sol ID, use $0.10 versus $0.26 on this turn, and read the GPT-5.6 Sol vs Terra vs Luna page for the older family.

Sol is the model to benchmark when the task is coding, tool use, or a multi-step agent and Astra's $10 / $50 is more than the task can return. Luna is the model when the same workflow is clerical and you can accept the smaller model. The cheapest models ledger ranks the rest of the catalog by agent-turn cost. Kimi K3 sits nearby at $3 / $15 if you want a non-OpenAI comparison at a similar sticker.

What OpenAI measured #

These figures are from OpenAI's 23 September 2026 announcement. They are vendor-run, at the effort level OpenAI names, against competitor scores OpenAI took from public reports. RouterPlex has not re-run them.

Luna's own scores, including DeepSWE 66.6% at max effort, are on the GPT-6 Luna pricing guide. Alignment notes, including coding-deception rates on deliberately hard cases, are in the GPT-6 Astra system card. Those evals are challenging cases, and OpenAI says they are not typical failure rates.

What the model card documents #

OpenAI's card says the Responses API is the path for built-in tools, and that Chat Completions supports function calling only with reasoning_effort set to none. On RouterPlex, call Sol through Chat Completions at https://api.routerplex.com/v1. Set that effort when the request includes tools. Hosted OpenAI tools on the card (web search, file search, code interpreter, computer use, and the rest) are an OpenAI Responses surface. Do not assume they are attached to this gateway route.

OpenAI also publishes standard rate limits on the card: tier 1 is 500 RPM and 500K TPM, and tier 5 is 15,000 RPM and 40M TPM. Those are OpenAI's limits. RouterPlex does not publish a separate per-model RPM table for this ID. The cap you set is the hard budget on the key.

Call GPT-6 Sol on RouterPlex #

One OpenAI-compatible key, prepaid, with a hard spend limit. The balance stops at $0. There is no per-token markup and no top-up fee.

curl https://api.routerplex.com/v1/chat/completions \
  -H "Authorization: Bearer $ROUTERPLEX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-6-sol",
    "messages": [
      {"role": "user", "content": "Name the riskiest assumption in this migration plan."}
    ]
  }'

import os
from openai import OpenAI
 
client = OpenAI(
    api_key=os.environ["ROUTERPLEX_API_KEY"],
    base_url="https://api.routerplex.com/v1",
)
 
response = client.chat.completions.create(
    model="gpt-6-sol",
    messages=[{"role": "user", "content": "Design a rollback-safe deployment plan."}],
)
 
print(response.choices[0].message.content)

Claude Code and the Anthropic SDK use ANTHROPIC_BASE_URL=https://api.routerplex.com and model gpt-6-sol. The Claude Code setup guide has the full configuration. Cursor can override the OpenAI base URL to https://api.routerplex.com/v1. Codex CLI can use a custom provider.

Live prices: GPT-6 Sol, GPT-6 Luna, GPT-6 Astra, and the OpenAI hub.

When Sol is the right ID #

• Coding agents and hard professional tasks where you want the GPT-6 tier and Astra's $0.50 turn is more than the task returns. Budget $2 / $10 and stay under 272K unless you mean to pay the cliff.
• A migration off GPT-5.6 Sol. The new ID is cheaper on this catalog ($0.10 versus $0.26 on the sample turn) and the input window is 922K rather than 258K. Re-run your own task set before you delete the old ID.
• High-volume extraction, classification, and triage. Use GPT-6 Luna at $0.10 / $0.50. Sol is twenty times the input price.
• The newest flagship. Use GPT-6 Astra. OpenAI still calls Astra the best model in the family.

Start a $5 RouterPlex test, put gpt-6-sol on a budgeted key, and send one live request. The same key reaches Luna, Astra, and the rest of the catalog. Gateways that fund credits with a percentage fee are compared on RouterPlex vs OpenRouter.