GPT-6.1 Sol API Pricing: Per-Token Costs and Workload Math

How much does the GPT-6.1 Sol API cost per 1M tokens?

GPT-6.1 Sol costs $2.00 per 1M input tokens, $0.10 per 1M cached input tokens, and $10.00 per 1M output tokens. The cached-input rate — 5% of the standard input price — is the cheapest OpenAI has published for a model in this class. Last verified: October 2026, against OpenAI's official GPT-6.1 Sol model page.

The Full Price Table

Straight from OpenAI's published rate card (standard tier, short context):

Token typePrice per 1M tokens
Input$2.00
Cached input$0.10
Cache writes$2.50
Output$10.00

Two details worth knowing. First, cached input is priced at exactly 5% of the uncached input rate, and cache writes cost 1.25x the input rate — so the first call that establishes a new cache prefix costs slightly more than an uncached call, and every reuse after that is nearly free. Second, this is the standard list price with no promo expiry attached — unlike GPT-5.6 Sol (promo "at least through November 21, 2026") and Gemini 4 Argon (intro rate through November 30, 2026), there is no published deadline after which GPT-6.1 Sol's price changes.

What Changed vs GPT-6 Sol

If you were already using GPT-6 Sol, here's the entire delta: nothing changed except cached input. Standard input and output are identical at $2/$10. Cached input dropped from $0.20 to $0.10 per 1M — halved, from 10% of the input rate to 5%. GPT-6.1 Sol launched September 29, 2026, and OpenAI positions it as near-Astra-class performance at a fifth of Astra's token prices (their claim, from the launch materials — treat vendor benchmarks accordingly).

The Workload Math: $97.68/Month

Headline rates don't pay invoices; workloads do. Run GPT-6.1 Sol through the reference workload — 1,100 tasks per month, each with 12K fresh input, 48K cached input, and 6K output tokens:

  • Fresh input: 12,000 x $2.00/1M = $0.0240 per task
  • Cached input: 48,000 x $0.10/1M = $0.0048 per task
  • Output: 6,000 x $10.00/1M = $0.0600 per task
  • Per task: $0.0888 → per month (x1,100): $97.68

Notice where the money goes: output is two-thirds of the per-task cost. The 48K cached tokens — four times the fresh input volume — cost a fifth as much as the fresh input. That's the 5% cache rate doing its job. On cache-heavy agentic workloads, GPT-6.1 Sol's cheaper cached reads are the whole story.

How It Compares on the Same Workload

Same 1,100-task workload, verified current prices:

ModelMonthly costvs GPT-6.1 Sol
GPT-6.1 Sol$97.68—
GPT-6 Sol$102.96+5.4%
Claude Opus 5.5$195.36+100%
Claude Sonnet 5$198.00+103%
Gemini 4 Argon (intro, thru Nov 30)$198.00+103%
GPT-5.6 Sol (promo, thru Nov 21)$205.92+111%
Gemini 4 Argon (from Dec 1)$396.00+305%

Three honest caveats. First, Sonnet 5 and Argon publish no cache rate, so their cached tokens are modeled at the full input rate — if your workload is cache-heavy, the gap to GPT-6.1 Sol is even wider than shown. Second, GPT-5.6 Sol's promo expires November 21 with no published successor price (pre-promo list was $5/$30, which would put it at $290.40/month). Third, these are standard-tier, short-context rates — the fine print below moves them.

The Fine Print That Moves Your Bill

  • Long context (>272K input tokens): the entire request reprices at 2x input and cache rates and 1.5x output. A 300K-token prompt doesn't get the $2/$10 headline — budget $4/$0.20/$15 for it.
  • Batch and Flex tiers: 50% off standard. If your workload tolerates delayed or off-peak processing, the reference workload drops to roughly $48.84/month.
  • Fast mode: 2x standard. Ultrafast mode: 6x standard. Only pay for speed when latency is the product.
  • Context window: 1,050,000 tokens, 128K max output. Knowledge cutoff April 2026.

FAQ

How much does GPT-6.1 Sol API cost per request?

$0.0888 on the reference workload (12K fresh input + 48K cached input + 6K output). A simple 1K-input / 1K-output request costs about $0.012.

Is GPT-6.1 Sol cheaper than GPT-6 Sol?

Slightly. Standard rates are identical ($2/$10); only cached input got cheaper ($0.10 vs $0.20). On the reference workload that's $97.68/month vs $102.96/month — about 5% less, all from the cache line.

Is GPT-6.1 Sol cheaper than Claude Opus 5.5?

Yes, roughly half the price: $97.68/month vs $195.36/month on the same workload. Opus 5.5 charges $4/$0.20/$20 — double the input and output rates.

Does GPT-6.1 Sol have a promo price that expires?

No. The $2/$10 rate is the standard list price with no published expiry — unlike GPT-5.6 Sol's promo (ends November 21, 2026) and Gemini 4 Argon's intro rate (doubles December 1, 2026). If you're choosing a model to build on past November, that stability is worth something the table doesn't show.

The Bottom Line

GPT-6.1 Sol is the cheapest flagship-class OpenAI model to actually run, and the reason is one number: $0.10 cached input. On workloads where the same context gets re-sent — agents, RAG, long sessions — it undercuts everything in its tier, and unlike its promo-priced rivals, the price isn't going anywhere on November 21 or December 1. Verify the rate card before you commit — prices move — but as of October 2026, this is the value pick.


 

Popular posts from this blog

GPT-5.6 API Pricing: The November 21 Deadline

Image
Last updated: October 2026 If you're running GPT-5.6 through the API right now, you're on a promotional price. And that promotion has an expiry date: November 21, 2026 . OpenAI's pricing page states the current rate is "available at least through November 21, 2026." Read that carefully. "At least through" — not "until." The promo could extend. It could also end exactly on that date. OpenAI has published no pricing for what comes after. This post lays out exactly what the pricing is today, what it costs on a realistic workload, and how to budget for the unknown that follows November 21. The Current Pricing, Exactly GPT-5.6 is available in several variants. The one most developers are using — the flagship reasoning model, referred to here as GPT-5.6 Sol — is currently priced at: Input: $4.00 per 1M tokens Cached input: $0.40 per 1M tokens Output: $20.00 per 1M tokens That is promotional pricing. Before the promo, announced August 21, ...
Read more

Argon vs Wonder: Which AI Should Developers Build On?

Image
Last updated: October 2026 Two flagship models, two very different price tags — and one of the prices is about to double. If you're choosing a foundation model to build on right now, the decision isn't just about capability. It's about timing. Here's the honest comparison: what each model costs, what the benchmarks actually say, and which one to pick depending on what you're optimizing for. The Pricing: Read the Dates Gemini 4 Argon is on introductory pricing: $2.00 per 1M input tokens / $10.00 per 1M output tokens . That intro rate runs through November 30, 2026 . From December 1 , the price doubles to $4.00 / $20.00 . Muse Wonder is priced at $4.00 / $20.00 per 1M tokens — no intro period, no scheduled change. So today, Argon costs half what Wonder does. On December 1, they cost exactly the same. That convergence matters more than it looks. Every cost comparison you run today has a 6-week expiry date. If you're doing the math — and you should be — do...
Read more

Prompt Caching: The 90% Discount You're Missing

Last updated: October 2026 There is a 90% discount sitting in your API bill, and most developers are either not using it or using it in a way that doesn't work. Prompt caching — paying a tenth of the input price for tokens the provider has already seen — is the highest-leverage cost optimization available on modern LLM APIs. Here's how it actually works and how to get the savings. How Prompt Caching Actually Works The idea is simple. When you send a request, the provider processes your input tokens through the model — that's the expensive computation you're paying for. But if your next request starts with the exact same tokens — the same system prompt, the same repo context, the same few-shot examples — the provider doesn't need to recompute all of that. It kept the intermediate results cached. So it charges you a fraction of the price for those repeated tokens. The key phrase is stable prefix . Caching works from the start of your prompt forward: the provider ...
Read more