As of 2026-11-09 · Sol promo $4/$20, official end date 2026-11-30

What does GPT-5.6 actually cost?
Tier prices, caching and real costs

As of 2026-11-09, GPT-5.6 Sol costs $4 per million input tokens and $20 per million output tokens, and OpenAI says this promotional pricing runs until 2026-11-30; OpenAI has not published the price after that date (official Sol model page and pricing page fetched 2026-11-09; the launch sticker was $5/$30). Terra $2/$12 per million tokens; real cost follows cache hits. Codex subscription usage is a separate meter: OpenAI’s @thsottiaux wrote on 2026-07-29 (fetched 2026-08-30) that OpenAI has not reduced usage on any subscription plans, that Sol was using Codex limits faster than expected, that usage limits were reset for all ChatGPT Work and Codex users, and that typical Sol use should last around 18% longer.

Prices change: on QCode, Claude, GPT and Chinese models are billed per token — live rates on /models.

#How the bill is counted #Cache hits #Sol promo $4/$20 to 11-30
Update 2026-07-30: Luna and Terra list prices cut

OpenAI cut Luna 80% and Terra 20% to $2/$12. Sol stayed at $5/$30 as of 2026-08-02, with Fast mode (the official pricing page fetched 2026-11-09 shows $4/$20, promotional through 2026-11-30). Track ongoing changes on the Luna price tracker.

Open the GPT-5.6 Luna price tracker

Official price table (per 1M tokens)

Cache writes bill at 1.25x the input rate, cache reads cost 90% less than uncached input. All three tiers share a 1,050,000-token API context and 128,000 max output; above 272K input the whole request bills at 2x input and 1.5x output. The Sol column shows current promo pricing (through 2026-11-30) -- the $5/$30 launch sticker appears in the examples below.

Tier Input Output Cache write (1.25x) Cache read (−90%)
gpt-5.6-sol $4 $20 $5.00 $0.40
gpt-5.6-terra $2 $12 $2.50 $0.20
gpt-5.6-luna $0.20 $1.20 $0.25 $0.02

The GA caching upgrades are worth designing for

GPT-5.6 ships more predictable prompt caching at GA: explicit cache breakpoints (you decide where the prefix splits) and a 30-minute minimum cache life. For agent workflows that means pinning system prompts and tool definitions as the cached prefix can cut long-session input costs by up to 90%.

Three reproducible cost examples

Line-by-line at official rates — swap in your own volumes to check.

Everyday agent session (Terra)

800K input + 60K output on Terra: 0.8 x $2 + 0.06 x $12 = $1.60 + $0.72 = about $2.32 (short-context list rates as of 2026-08-02). If 80% of the input is a cached read ($0.20/M), the input part is 0.16 x $2 + 0.64 x $0.20 = $0.448, so about $1.17 total.

Hard task (Sol, high cache hit)

500K input (400K of it cached reads) + 50K output at Sol's current price per the official pricing page fetched 2026-11-09 ($4 input, $0.40 cached read, $20 output; promotional through 2026-11-30): 0.1 x $4 + 0.4 x $0.40 + 0.05 x $20 = $0.40 + $0.16 + $1.00 = about $1.56 (about $2.20 at the $5/$30 launch price). Caching brings a Sol deep-dive close to Terra's uncached price for the same volume (about $1.60).

Bulk light tasks

1,000 calls at 10K input + 1K output each (10M input + 1M output): converted on the official price table (checked 2026-11-09), that is about $60 on Sol at the current $4/$20 and about $32 on Terra at the current $2/$12. On QCode, run this kind of high-volume repeat work on gpt-5.6-terra (QCode does not currently support gpt-6-luna or gpt-5.6-luna).

How the prices compare

On current official list prices (checked 2026-11-09): Sol is $4 input and $20 output, promotional through 2026-11-30, with the price after that date not published by OpenAI; Terra is $2 input and $12 output and works as the everyday default tier. For high-volume, low-cost work on QCode, use gpt-5.6-terra (QCode does not currently support gpt-6-luna or gpt-5.6-luna).

Prices and usage on QCode

QCode is billed per token; per-model rates are listed on /models. gpt-5.6-sol and gpt-5.6-terra are available on QCode and share one API key and one quota with the other GPT models on QCode, Claude and the Chinese models, with self-service top-ups and activation. For more GPT options you can also use gpt-6.1-sol or gpt-6-sol (QCode does not currently support gpt-6-luna or gpt-5.6-luna). A request with more than 272K input tokens bills the whole request at the long-context rate: input and cache-read prices x2, output x1.5, in line with OpenAI's rule. Payment methods: Alipay, WeChat Pay, credit card (Visa / Mastercard), Apple Pay, Google Pay, cryptocurrency (multiple coins and networks). Not ready for a monthly plan? Pay-as-you-go Packs give you a one-time quota to use within 7 days (see /pay-as-you-go); if you already have a Personal Plan, add a booster pack when you need more quota.

Pricing FAQ

How exactly is caching billed?

Cache writes bill at 1.25x the uncached input rate. As of 2026-11-09, Sol's official model page lists $4 input, $0.40 cached input and $20 output per 1M tokens, which converts to a $5.00 cache write on the official price table; the Terra page lists $2 input, $0.20 cached input and $12 output, giving a $2.50 cache write under the same rule; Sol's promotional pricing runs through 2026-11-30. The official pricing page (fetched 2026-11-09) splits cache writes into two columns: the 30m write is the $5.00 (Sol) / $2.50 (Terra) above, and the 12h write is $10.00 (Sol) / $5.00 (Terra), both per 1M tokens at short-context rates. The official prompt-caching guide states that both implicit and explicit caching are supported, with explicit caching giving you more control over which context is written to cache, and that a cached prefix remains eligible for reuse for 30 minutes after its most recent write or reuse. Historically: as of 2026-08-02 Sol was still listed at $5/$30, where the same rule gave a $6.25 cache write and a $0.50 cache read.

Is long context billed extra?

There is a long-context tier. OpenAI's Sol model page says that once input exceeds 272K, the whole request bills at 2x input and 1.5x output ("Prompts with >272K input tokens are priced at 2x input and 1.5x output for the full request."). Note 272K is not a ceiling on what you can send -- the window stays 1,050,000 -- it is a price cliff. The lever is still cache hits on a stable prefix, not truncating context. On the official pricing page (fetched 2026-11-09) the long-context rates are $8.00 input, $30.00 output and $0.80 cached input for Sol, and $4.00, $18.00 and $0.40 for Terra (per 1M tokens). QCode bills by this same rule.

Does switching tiers affect billing or quota?

No. The tiers are just different model ids, each billed at its own rate against the same quota, so you can mix them per request in one project. As of 2026-11-09 gpt-5.6-sol and gpt-5.6-terra are available on QCode — say Sol to plan, Terra to execute, and Terra again for high-volume repeat runs. QCode does not currently support gpt-6-luna or gpt-5.6-luna; please use another GPT model.

When does the GPT-5.6 Sol promo price end, and what will it cost after that?

The promotional price runs until 2026-11-30, and OpenAI has not published the price after that date. The official Sol model page (fetched 2026-11-09) states GPT-5.6 Sol’s promotional pricing is available until November 30, 2026; the current rate is $4 input, $0.40 cached input and $20 output per 1M tokens, which OpenAI describes as a 20% reduction in input pricing and a 33% reduction in output pricing. The launch sticker was $5/$30 (still the list price as of 2026-08-02); Terra is $2/$12. QCode is billed per token; per-model rates are listed on /models. gpt-5.6-sol and gpt-5.6-terra are available on QCode; QCode does not currently support gpt-6-luna or gpt-5.6-luna, so use gpt-6.1-sol, gpt-6-sol or gpt-5.6-terra instead.

Use GPT-5.6 on QCode, billed per token

GPT-5.6 Sol and Terra are included in a QCode monthly plan, billed per token with per-model prices on /models — from $9/mo, month to month, cancel whenever. On one key, switching between gpt-5.6-sol, gpt-5.6-terra, gpt-6.1-sol and gpt-6-sol is just a model id change; QCode does not currently support gpt-6-luna or gpt-5.6-luna.

Try first, then decide

Not sure which tier? Start with Starter ($9.00/mo) and upgrade when you're happy — the unused value of the old plan goes back to your balance.