Education

OpenAI API Pricing 2026: GPT-6.2 Sol, Astra, Terra, Luna

OpenAI API pricing for every current GPT model, including the new GPT-6.2 Sol and Astra, compared direct, through Azure, Bedrock, OpenRouter and Requesty, and on SayGM, where most GPT models are 48.5% below OpenAI's standard rate.

BSBrittany Seales· Marketing5 min read
OpenAI Pricing Compared

Updated 19 November 2026. GPT-6.2 Sol, OpenAI's newest Sol model, costs $2.00 per million input tokens and $10.00 output on OpenAI. SayGM serves it at $1.75 and $8.76, 12.5% below. On most of the current GPT line SayGM is 48.5% below OpenAI's standard rate: GPT-6.1 Sol is $1.03 and $5.15, so 100 million tokens cost $185 on SayGM against $360 direct and about $380 through Requesty or OpenRouter, which both add a fee on top of OpenAI's price.

OpenAI added GPT-6.2 Sol and GPT-6.2 Astra to its API around 17 November, and GPT-6.1 Terra in late October, so most pricing tables online are already out of date. Below is what every current GPT model costs on each route, what OpenAI's speed tiers and long prompts add, and which model is worth paying for.

Table of Contents

OpenAI API pricing at a glance

OpenAI API pricing by route on 19 November 2026: SayGM is the cheapest way to buy GPT-6.2 Sol, GPT-6.1 Sol and GPT-6.2 Astra

SayGM is the cheapest way to buy every GPT model in this post at standard speed, the speed every other route in this post sells by default. Most are 48.5% below OpenAI's standard rate, from GPT-6 Luna at $0.05 and $0.26 to GPT-6 Astra at $5.15 and $25.75. Four are on a smaller discount: GPT-6.2 Sol, GPT-6.2 Astra and o3 at 12.5% below OpenAI, and GPT-6.1 Terra at 43.4% below. The 100-million-token figures throughout this post assume an 80/20 input-to-output mix.

The exception is OpenAI's own Flex and Batch tiers, which charge half its standard rate. That's slightly below SayGM's standard price on the 48.5% models, and well below it on GPT-6.2 Sol, GPT-6.2 Astra and o3. Flex responses are deliberately slower and can be refused when OpenAI is short on capacity, and Batch returns results later, so both suit evaluations and background jobs rather than anything a user is waiting on. The comparisons in this post are all at standard speed.

GPT model prices: OpenAI vs SayGM vs Requesty vs OpenRouter

OpenAI's standard rates come from OpenAI's pricing page. Requesty passes OpenAI's price through with a 5% markup, and OpenRouter with a 5.5% fee when you buy credit on its Standard plan. SayGM's rates are from its live catalog. All checked 19 November 2026.

Each cell shows input / output per million tokens, then the cost of 100 million tokens at an 80/20 mix.

ModelOpenAI directSayGMRequestyOpenRouter
GPT-6.2 Sol$2.00 / $10.00 · $360$1.75 / $8.76 · $315$2.10 / $10.50 · $378$2.11 / $10.55 · $380
GPT-6.2 Astra$8.00 / $40.00 · $1,440$7.00 / $35.02 · $1,261$8.40 / $42.00 · $1,512$8.44 / $42.20 · $1,519
GPT-6.1 Sol$2.00 / $10.00 · $360$1.03 / $5.15 · $185$2.10 / $10.50 · $378$2.11 / $10.55 · $380
GPT-6.1 Terra$1.00 / $5.00 · $180$0.57 / $2.83 · $102$1.05 / $5.25 · $189$1.06 / $5.28 · $190
GPT-6 Astra$10.00 / $50.00 · $1,800$5.15 / $25.75 · $927$10.50 / $52.50 · $1,890$10.55 / $52.75 · $1,899
GPT-6 Sol$2.00 / $10.00 · $360$1.03 / $5.15 · $185$2.10 / $10.50 · $378$2.11 / $10.55 · $380
GPT-6 Luna$0.10 / $0.50 · $18$0.05 / $0.26 · $9$0.11 / $0.53 · $19$0.11 / $0.53 · $19
GPT-5.6 Sol$4.00 / $20.00 · $720$2.06 / $10.30 · $371$4.20 / $21.00 · $756$4.22 / $21.10 · $760
GPT-5.6 Terra$2.00 / $12.00 · $400$1.03 / $6.18 · $206$2.10 / $12.60 · $420$2.11 / $12.66 · $422
GPT-5.6 Luna$0.20 / $1.20 · $40$0.10 / $0.62 · $21$0.21 / $1.26 · $42$0.21 / $1.27 · $42
GPT-5.5$5.00 / $30.00 · $1,000$2.58 / $15.45 · $515$5.25 / $31.50 · $1,050$5.28 / $31.65 · $1,055
GPT-5.4$2.50 / $15.00 · $500$1.29 / $7.73 · $258$2.62 / $15.75 · $525$2.64 / $15.82 · $528
GPT-5.4 mini$0.75 / $4.50 · $150$0.39 / $2.32 · $77$0.79 / $4.73 · $158$0.79 / $4.75 · $158
GPT-5.4 nano$0.20 / $1.25 · $41$0.10 / $0.64 · $21$0.21 / $1.31 · $43$0.21 / $1.32 · $43
o3$2.00 / $8.00 · $320$1.75 / $7.00 · $280$2.10 / $8.40 · $336$2.11 / $8.44 · $338

OpenAI's pricing page also lists promotional pricing on GPT-5.6 Sol until 30 November 2026; the table shows its standard rate.

GPT-6.2 Sol and Astra: what's new

GPT-6.2 Sol and GPT-6.2 Astra reached the API around 17 November. GPT-6.2 Sol keeps 6.1 Sol's headline rate of $2.00 and $10.00, with cached input at $0.10. GPT-6.2 Astra is $8.00 and $40.00, which makes it cheaper than GPT-6 Astra's $10.00 and $50.00 on OpenAI's own price list.

On SayGM the picture is different, because the new models carry a smaller discount than the rest of the range for now. GPT-6.2 Sol is $1.75 and $8.76, 12.5% below OpenAI, while GPT-6.1 Sol is $1.03 and $5.15, 48.5% below. Per 100 million tokens that's $315 against $185. If 6.1 Sol does the job, it's the cheaper Sol by a wide margin today. The same applies to Astra: GPT-6 Astra is $927 per 100 million tokens on SayGM, against $1,261 for GPT-6.2 Astra.

OpenAI hasn't published benchmark comparisons for 6.2 against 6.1 on the pages checked for this post, so test both on your own prompts before switching.

GPT-6 family pricing: Astra, Sol, Terra or Luna

The GPT-6 family now spans four tiers, and picking the right one matters more than picking the right route.

GPT-6 Astra and GPT-6.2 Astra are OpenAI's most capable models, at $10.00 and $50.00 for GPT-6 Astra and $8.00 and $40.00 for 6.2 Astra. If a Sol model handles the task, Astra costs four to five times as much.

GPT-6.1 Sol and GPT-6.2 Sol are the default for most production work, especially agents and coding. GPT-6.1 Sol bills cached input at $0.10 per million, half GPT-6 Sol's $0.20, which matters for agents and coding tools that resend the same system prompt and files on every turn. There's no reason to start new work on GPT-6 Sol.

GPT-6.1 Terra sits between Sol and Luna at $1.00 and $5.00, half Sol's price, for work that needs more than Luna but not a full Sol model. On SayGM it's $0.57 and $2.83, about $102 per 100 million tokens.

GPT-6 Luna is $0.10 and $0.50, built for fast, high-volume tasks: classification, extraction, routing and short replies. It's the cheapest current GPT model by a wide margin, at $9 per 100 million tokens on SayGM.

What long prompts and speed tiers add to the bill

Two things push an OpenAI bill above the headline rate.

Prompts over 272K tokens. The GPT-6 models accept up to about 1M tokens of context, but past 272K the rate steps up to double the input price and one and a half times the output. For GPT-6.1 Sol that's $4.00 and $15.00 on OpenAI, and $2.06 and $7.73 on SayGM. For GPT-6.2 Sol it's the same $4.00 and $15.00 on OpenAI, and $3.50 and $13.13 on SayGM.

Speed tiers. OpenAI sells each model at several speeds. Flex and Batch are half the standard rate, for work that can wait. Standard is the rate everyone quotes. Fast costs twice as much for lower latency, and GPT-6 Astra and GPT-6.1 Sol also have an Ultrafast tier at six times standard; for GPT-6.2 Sol and Astra, Ultrafast is three times standard until 28 February 2027 and six times from 1 March. OpenAI's pricing page also lists a 10% uplift for data residency on models released on or after 5 March 2026.

GPT-5.6, GPT-5.5, GPT-5.4 and o3 pricing

Plenty of production traffic still runs on older models, and they're all in the table above. A few are worth a closer look.

GPT-5.5 at $5.00 and $30.00 costs nearly three times as much as GPT-6.1 Sol per 100 million tokens, for an older model. If you're still on it, moving is the quickest cut available. GPT-5.6 Sol at $4.00 and $20.00 costs twice as much as GPT-6.1 Sol.

o3 pricing is $2.00 and $8.00 on OpenAI, or $1.75 and $7.00 on SayGM, 12.5% below. It remains a capable reasoning model, though GPT-6.1 Sol now costs less on SayGM for most work.

o4-mini no longer appears on OpenAI's pricing page or SayGM's catalog, so it has been dropped from the table.

Where to buy GPT models

GPT models are closed, so every route ends at OpenAI's models. What changes is the price, the extras around it, and who sits in the path. Prices below are for GPT-6.1 Sol per million tokens, input / output, unless stated.

OpenAI direct

OpenAI Home Page

$2.00 / $10.00. Buying direct is the only way to get every speed tier: Flex and Batch at half price for work that can wait, Standard, Fast at twice the price, and Ultrafast on selected models. The trade-off is that you pay the list rate, with a 10% uplift if you need data residency.

Azure

Azure AI Home Page

Azure lists GPT-6.1 Sol on global and data-zone endpoints. Its pricing page loads the rates by region, and they weren't readable when we checked on 19 November, so check the page for your region. Azure suits teams that already buy cloud from Microsoft and want GPT billed and governed alongside the rest of it.

Amazon Bedrock

Amazon Bedrock Home Page

Bedrock now lists the full GPT-6 family, including GPT-6.2 Sol and Astra, at Standard, Priority, Flex and Batch speeds. It's the route for AWS-first teams who want GPT next to their other AWS services and under their existing AWS agreement. Check its pricing page for your region's rates.

OpenRouter

OpenRouter Home Page

$2.11 / $10.55 once its 5.5% fee on buying credit is added, or 8% on its Business plan. OpenRouter passes through OpenAI's own endpoints, and adds routing on top: automatic failover when one endpoint errors, and one key for hundreds of models. We break it down further in our OpenRouter alternatives comparison.

Requesty

Requesty Home Page

$2.10 / $10.50 with its 5% markup. Requesty includes routing, caching and EU data residency on every plan, and supports bringing your own provider keys.

SayGM

SayGM Home Page

$1.03 / $5.15, 48.5% below OpenAI's standard rate, with no subscription and no credit fee. Long prompts over 272K tokens are $2.06 / $7.73 against OpenAI's $4.00 / $15.00. SayGM also carries Claude, Gemini and open-weight models behind the same key, and switching is a base URL change:

python
import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.saygm.com/v1",
    api_key=os.environ["GM_API_KEY"],
)

response = client.chat.completions.create(
    model="gpt-6.1-sol",
    messages=[{"role": "user", "content": "Review this pull request for bugs."}],
)
print(response.choices[0].message.content)

The GPT-6.1 Sol model page shows the live rate.

What OpenAI sees when you use SayGM

OpenAI still receives the prompt on every route, because it runs the model. On SayGM, requests use anonymous routing, so OpenAI gets the request but not who sent it, and the gateway in between runs inside hardware-attested confidential compute.

You can also strip sensitive data before OpenAI sees it. SayGM's guardrails are optional, set per API key, and redact personal data, secrets and credentials inside the gateway's enclave before the request leaves. They're pattern-based, so treat them as one layer of protection rather than a complete one.

What changed since October

This post first went up on 7 October 2026. Here's what has moved since, checked on 19 November 2026.

  • SayGM's GPT discount rose from 22% to 48.5% on most models. GPT-6.1 Sol fell from $1.56 and $7.80 to $1.03 and $5.15, so 100 million tokens went from $281 to $185.
  • GPT-6.2 Sol and GPT-6.2 Astra reached the API around 17 November, at $2.00 / $10.00 and $8.00 / $40.00 on OpenAI. SayGM carries both at 12.5% below.
  • GPT-6.1 Terra joined the range at $1.00 / $5.00, between Sol and Luna.
  • OpenRouter's 50% promotion on GPT-5.6 Sol has ended; it's back to OpenAI's standard rate plus the credit fee. OpenAI's own GPT-5.6 Sol promotion runs until 30 November.
  • OpenAI's speed tiers are now Flex, Batch, Standard, Fast and Ultrafast. Priority processing was renamed Fast on 30 July, and Ultrafast is available on its top models.
  • o4-mini is gone from OpenAI's pricing page and SayGM's catalog. GPT-5.6 Terra, GPT-5.6 Luna and GPT-5.4 nano are added to the table.

OpenAI API cost questions

How is OpenAI API cost calculated?

Per token, with separate rates for input, cached input and output. Cached input is a tenth of the standard input rate on most current models, and a twentieth on GPT-6.1 Sol and GPT-6.2 Sol, which is why reusing the same prompt prefix across calls lowers the bill.

What is the cheapest OpenAI model?

GPT-6 Luna, at $0.10 and $0.50 on OpenAI and $0.05 and $0.26 on SayGM, as of 19 November 2026.

How much does GPT-6.2 Sol cost?

$2.00 per million input tokens and $10.00 output on OpenAI, the same headline rate as GPT-6.1 Sol. On SayGM it's $1.75 and $8.76, while GPT-6.1 Sol is $1.03 and $5.15, so 6.1 Sol is the cheaper Sol on SayGM today.

Is GPT API pricing cheaper through a gateway?

It depends on the gateway. OpenRouter and Requesty add a fee on top of OpenAI's price, 5.5% and 5% respectively. SayGM is below OpenAI's price on every GPT model at standard speed: 48.5% below on most, and 12.5% below on GPT-6.2 Sol, GPT-6.2 Astra and o3.

About SayGM

SayGM is a drop-in inference gateway for teams who don't want to just take a company's word that their prompts are private. Route to a confidential model and your prompt is sealed inside hardware-verified execution, where not even SayGM can see it. That's not a policy, it's provable. Swap in your existing OpenAI, Anthropic, or Gemini code and you're covered in minutes, at transparent, published rates with no hidden markup.

Say gm to AI at saygm.com.

Website | Twitter | Discord | Blog | Medium | Docs

  • ChatGPT
  • GPT
  • OpenAI
  • Pricing
BS

Brittany SealesMarketing

Saying gm to marketing (and AI)

X ↗