GPT-5.6 API Pricing: Sol vs Terra vs Luna

GPT-5.6 Sol API pricing on RouterPlex is $5 per 1M input tokens and $30 per 1M output tokens. Terra is $2 / $12. Luna is $0.20 / $1.20. Those are the catalog rates this route is charged, after the 30 July 2026 cut to Terra and Luna.

OpenAI currently lists Sol at a promotional $4 / $20, at least through 21 November 2026. RouterPlex does not bill that promo. If you are costing a Sol workload from openai.com/pricing, double-check the live GPT-5.6 Sol page before you lock a budget.

The three API model IDs are gpt-5.6-sol, gpt-5.6-terra and gpt-5.6-luna. All three are live on RouterPlex with image input, 128K maximum output and a 258K input limit on the current master route.

OpenAI's newer flagship is GPT-6 Astra at $10 / $50, with a 272K cliff. It sits above this family, not inside it. On 23 September 2026 OpenAI also shipped GPT-6 Sol at $2 / $10 and GPT-6 Luna at $0.10 / $0.50. This page stays on the GPT-5.6 catalog rates.

GPT-5.6 API price comparison #

OpenAI's public Sol promo ($4 / $20) is not in this table because it is not what this route bills. Cached input is one tenth of standard input for these models on OpenAI's sheet. Output remains the largest cost category, so concise completion limits and clear stopping conditions matter more than small prompt changes in many workloads.

On a 40,000-token prompt returning 2,000 tokens, Sol costs $0.26, Terra $0.104, Luna $0.0104. GPT-6 Astra on the same turn is $0.50.

GPT-5.6 Sol vs Terra vs Luna #

Choose GPT-5.6 Sol for the hardest 5.6 work

Sol has the highest GPT-5.6 price and is positioned for complex coding, long-document reasoning, difficult analysis and image understanding. Use it when a better 5.6 answer is worth more than token savings, and when you do not need Astra's newer weights or 922K input window.

Live pricing: GPT-5.6 Sol API.

Choose GPT-5.6 Terra for a balanced default

Terra costs two fifths of Sol on both standard input and output ($2 / $12 against $5 / $30). That makes it the practical starting point for everyday coding assistance, document work, customer-facing AI features and agents whose tasks are important but not consistently frontier-hard.

Live pricing: GPT-5.6 Terra API.

Choose GPT-5.6 Luna for volume

Luna costs one twenty-fifth of Sol on input and output ($0.20 / $1.20). It fits classification, extraction, straightforward Q&A, document triage and high-frequency automation where a lower per-request cost creates more value than maximum capability.

Live pricing: GPT-5.6 Luna API.

Use GPT-5.6 through one OpenAI-compatible endpoint #

The model string is the only part that changes:

curl https://api.routerplex.com/v1/chat/completions \
  -H "Authorization: Bearer $ROUTERPLEX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.6-terra",
    "messages": [
      {"role": "user", "content": "Find the failure modes in this deployment plan."}
    ]
  }'

To compare all three, run the same representative prompt set with a separate hard budget for each model. Record output quality, latency, uncached input, cached input and output cost instead of judging from one demo prompt. Add gpt-6-astra to that matrix if the task might justify the flagship rate.

Context and vision behavior #

OpenAI's direct GPT-5.6 model pages list a 1,050,000-token context window and 128,000 maximum output tokens. RouterPlex currently advertises 258K input because that is the limit supplied by its master route. The smaller published RouterPlex limit is intentional: customers should plan against the route they can actually call, not a larger direct-vendor specification.

Sol, Terra and Luna all accept images. That makes the same endpoint useful for screenshots, diagrams, scanned documents and mixed text-image tasks. The output remains text.

A simple selection rule #

Start with Terra. Move to Sol only when evaluation shows a meaningful quality improvement on the task that pays for the higher rate. Move to Luna when Terra succeeds reliably and throughput or unit cost matters more. Move to GPT-6 Astra only when the newer flagship is the thing you are actually buying — and stay under 272K unless you mean to pay that cliff.

This approach is more reliable than choosing from benchmark headlines or from OpenAI's promotional Sol floor. Your prompt structure, tools, repository and error tolerance determine which price-quality point is best. The number that matters is the catalog rate, not the marketing page.

Start a $5 RouterPlex test, set a hard budget on the key, and compare GPT-5.6 with GPT-6 Astra, Kimi K3 and Claude Fable 5. The rest of the OpenAI family is on the OpenAI pricing hub. Ranked by agent-turn cost, the full catalog is on the cheapest models ledger.