OpenAI-Compatible GPT Models Pricing
Updated · official prices verified at the source
faborapi gives access to GPT models through one OpenAI-compatible API and a prepaid balance. With the $29 → $320 credit pack, real cash cost = credit rate × 0.090625, so compared GPT models cost 86% to 91% less than OpenAI's official rates. For example, GPT-6 Astra costs $0.906 / $4.53 instead of $10.00 / $50.00 per 1M input/output tokens.
- GPT models
- 7GPT models
- max saving
- 91%max saving
- real cost factor
- ×0.091real cost factor
GPT price comparison per 1M tokens
Pay $29 for $320 API credits (91% pack discount). Real cost = credit rate × 0.090625. Savings compare that real cost with OpenAI's official USD rate for the exact same model ID.
| Model | Official price (input / output) | faborapi credit rate | Real cost with $29 pack | Real savings | Official source verified |
|---|---|---|---|---|---|
GPT-6 Astragpt-6-astra | $10.00 / $50.00 | $10.00 / $50.00 | $0.906 / $4.53 | 91% less | OpenAI pricing (opens in a new tab)Checked |
GPT-5.6 Solgpt-5.6-sol | $4.00 / $20.00 | $5.00 / $30.00 | $0.453 / $2.72 | 89% / 86% less | OpenAI pricing (opens in a new tab)Checked |
GPT-6.1 Solgpt-6.1-sol | $2.00 / $10.00 | $2.00 / $10.00 | $0.181 / $0.906 | 91% less | OpenAI pricing (opens in a new tab)Checked |
GPT-5.6 Terragpt-5.6-terra | $2.00 / $12.00 | $2.00 / $12.00 | $0.181 / $1.09 | 91% less | OpenAI pricing (opens in a new tab)Checked |
GPT-5.6 Lunagpt-5.6-luna | $0.20 / $1.20 | $0.20 / $1.20 | $0.018 / $0.109 | 91% less | OpenAI pricing (opens in a new tab)Checked |
GPT-6 Lunagpt-6-luna | $0.10 / $0.50 | $0.10 / $0.50 | $0.009 / $0.045 | 91% less | OpenAI pricing (opens in a new tab)Checked |
GPT-6gpt-6 | No exact official ID match | $5.00 / $10.00 | $0.453 / $0.906 | Not comparable | OpenAI pricing (opens in a new tab)ID not listed, checked |
Official rates are OpenAI's Standard short-context prices; Batch, Flex, Fast and long-context prices differ. “No exact official ID match” means OpenAI's pricing page lists no price for that exact model ID.
GPT-4o, GPT-4o mini and o1 are not in the faborapi catalog. Only the GPT model IDs listed in this table can be called.
faborapi rates are read live from the catalog by exact model ID. Cache and long-context rates differ; see the documentation. Failed requests are not billed.
Credit packs for GPT: minimum top-up $10
Credit packs range from the Nano Pack ($10 for $100 of credit, the minimum top-up) to the Gold Pack ($99 for $1100). Every pack works with all GPT models above and with every other model in the catalog. Savings figures use the ×0.090625 factor shared by the Starter packs. Bronze and Gold lower it to 0.0891 and 0.09 respectively, so their real token cost is lower still.
Choose a pack
You pay
$29
API credit added
$320
- Cost per $100 of credit
- $9.06
- Real cost factor
- × 0.0906
- Below credit face value
- 90.94%
- Models unlocked
- Every model in the catalog
Prepaid, no subscription. Same models and token rates with every pack.
| Pack | You pay | API credit added | Cost per $100 of credit | Real cost factor | Below credit face value |
|---|---|---|---|---|---|
| Nano PackMinimum top-up | $10 | $100 | $10.00 | × 0.1 | 90% |
| Mini Pack | $15 | $160 | $9.38 | × 0.0938 | 90.63% |
| Starter PackReference pack | $29 | $320 | $9.06 | × 0.0906 | 90.94% |
| Bronze Pack | $49 | $550 | $8.91 | × 0.0891 | 91.09% |
| Silver Pack | $89 | $930 | $9.57 | × 0.0957 | 90.43% |
| Gold Pack | $99 | $1100 | $9.00 | × 0.09 | 91% |
Why use faborapi for GPT?
One key, 25 models
Your sk-fabor- key covers Claude, GPT, Gemini and Grok. Switch models by changing the model string.
Prepaid, no subscription
Each successful request is deducted from one prepaid balance and itemized on your Usage page.
Drop-in compatible
Point an existing OpenAI SDK at https://faborapi.com/v1 and keep the Chat Completions request format.
GPT pricing FAQ
How is the faborapi discount on GPT API pricing calculated?
faborapi deducts credit at its listed rate per 1M tokens, and the Starter pack costs $29 for $320 of credit, so real cash cost = credit rate × 0.090625. OpenAI lists GPT-6 Astra (gpt-6-astra) at $10.00 / $50.00 per 1M input/output tokens; faborapi deducts $10.00 / $50.00 in credit, which costs $0.906 / $4.53 with the pack, 91% less. Official prices were checked on 29 October 2026.
Can I call GPT models with the OpenAI SDK?
Yes. faborapi serves GPT models through the OpenAI-compatible POST https://faborapi.com/v1/chat/completions endpoint. Set base_url to https://faborapi.com/v1, use your sk-fabor- API key and an exact model ID such as gpt-6-astra. faborapi implements Chat Completions and the stateless OpenAI Responses API (POST /v1/responses with store=false); Assistants and Realtime APIs are not available.
How does the prepaid balance work for GPT requests?
Every successful GPT request deducts its token cost from one prepaid USD balance shared by all faborapi models. Failed requests are not billed, and a request is refused with HTTP 402 when the balance is 0 or lower. New accounts receive $0.50 of welcome credit. Credit packs range from the Nano Pack ($10 for $100 of credit, the minimum top-up) to the Gold Pack ($99 for $1100). The $29 Starter pack adds $320 of credit. Every pack is paid by bank card, Manza in Morocco or Mobile Money in supported African countries.
Start with $320 of GPT credit for $29
Create an account, top up, and make your first GPT API call in under two minutes.
How it works in Cursor & VS Code
Connect your editor to the GPT models currently listed in faborapi without rebuilding your application. Cursor supports custom OpenAI settings; in VS Code, use an extension that accepts an OpenAI-compatible endpoint, such as Continue. These instructions configure chat requests, not every proprietary editor feature.
- 1
Prepare your account. Sign up, confirm your email, and create an API key. Check your prepaid balance and copy an exact model ID from the live catalog; a model mentioned in a comparison may not currently be available.
- 2
Set the connection. In Cursor’s model settings, enter your faborapi key and override the OpenAI base URL. In Continue, choose the OpenAI provider and set the same address: https://faborapi.com/v1. Add the selected model ID explicitly. Menu names vary by version.
- 3
Test before switching. Send a short prompt, confirm the response, and inspect Usage for token deductions. Keep production keys in environment variables, never in committed code. Check extension support before enabling streaming or tools.
quickstart.py
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["FABORAPI_API_KEY"],
base_url="https://faborapi.com/v1", # Replace the default endpoint
)
response = client.chat.completions.create(
model="gpt-6-astra",
messages=[{"role": "user", "content": "Explain this function briefly."}],
)
print(response.choices[0].message.content)