OpenAI / ChatGPT API Pricing — KeyoAPI Rates & Costs

Searching ChatGPT API pricing, OpenAI API pricing, or OpenAI API costs? This page compares indicative KeyoAPI sell rates for GPT-class IDs — including OpenAI-compatible gpt-6-astra — with typical public OpenAI list bands, then shows a simple token-cost planning method on one OpenAI-compatible base URL and prepaid wallet.

OpenAI / ChatGPT API Pricing is a model-family pricing lane on KeyoAPI — a prepaid AI API gateway where one Bearer key and one wallet ($1 = 500,000 quota units) cover chat, speech, vision, OCR and video models.

Price comparison table

Indicative figures for planning. Confirm live Keyo sell rates on interactive Model Square or the static pricing list.

Prototype on gift-credit models, then promote to GPT-class

To prototype on the same key, use free catalog IDs on the gift-credit models page — gift-credit IDs that bill at dedicated gift-credit rates from your one-time $10 new-user credit (fair-use rate limits). When quality gates pass, switch model= to gpt-6-astra, gpt-5.6-luna, deepseek-v4.1-flash, or kimi-k3 without changing base URL or auth. Guides: the gpt-6-astra model page, the deepseek-v4.1-flash model page, the kimi-k3 model page.

Cheapest OpenAI-class API on Keyo

If you searched for a cheap OpenAI API, two different questions are hiding in there. (1) The cheapest GPT-5-class lane on Keyo is gpt-5.6-luna at ~$0.10 / $0.62 per 1M in/out — GPT-5-class twins under the same OpenAI-compatible base URL. (2) The cheapest GPT-6-class lane on that same key is gpt-6-luna at ~$0.03 / $0.15 per 1M in/out; deepseek-v4.1-flash-1017 (~$0.02 / $0.06) is also available when you want high-volume entry lanes in other families.

Both ride one key, one base URL and one prepaid wallet, so the cheapest way to test OpenAI-shaped code is usually to prototype on a gift-credit ID from the gift-credit models page, then move only the traffic that needs GPT-class quality onto luna.

Cost math: 50M input + 10M output tokens/month on luna ≈ $5.00 in + $6.20 out ≈ $11.20. The same mix on gpt-6-astra at ~$1.84 / $9.19 ≈ $92.00 + $91.90 ≈ $183.90. Confirm live Keyo rates on the full pricing list before you budget.

Rates and how to call

Teams that search ChatGPT API pricing, OpenAI API pricing, or OpenAI API costs usually need GPT-class chat inside an OpenAI SDK with clearer unit economics — or one prepaid balance that also covers Claude, DeepSeek, Whisper, OCR, and TTS. Official OpenAI / ChatGPT surfaces win on brand trust and first-party product UI. KeyoAPI is for when you want that stack without five invoices and without rewriting clients when you swap models.

Keyo rates in the table are indicative sell prices from the live Keyo catalog. On Keyo, gpt-6-astra is a Keyo-hosted, OpenAI-compatible GPT-class model ID (confirm live rates on the gpt-6-astra model page). Always confirm interactive try/buy under the pricing page (Model Square) for gpt-6-astra, gpt-5.6-luna, gpt-5.6-terra, and gpt-5.6-sol before contracting volume. OpenAI list bands, cached-input discounts, and model renames change — treat official figures as planning context only. Keyo GPT-5.6 IDs remain GPT-class twins on an OpenAI-compatible API, not OpenAI first-party model names.

Worked planning example (indicative only): suppose a batch path burns 50M input tokens and 10M output tokens per month. At Luna ~$0.10 / $0.62 per 1M in/out that is roughly $5 input + $6.20 output ≈ $11.20 before retries and overhead. At Terra ~$0.37 / $2.21 the same mix is roughly $18.50 + $22.10 ≈ $40.60. Put high-volume classification on Luna and keep product chat on Terra so the expensive model only sees the traffic that needs it. Re-run the math on the pricing page before you budget.

Integration steps: (1) create an account at the sign-up page and top up Wallet; (2) create an API key in console; (3) set base URL https://www.keyoapi.xyz/v1 (OPENAI_BASE_URL or sdk base_url); (4) call POST /v1/chat/completions with model gpt-6-astra, gpt-5.6-luna, gpt-5.6-terra, or gpt-5.6-sol; (5) try/buy interactively on the pricing page (Model Square) — there is no separate upload playground. Longer guides: the gpt-6-astra model page, the gpt-5.6-luna model page, the gpt-5.6-terra model page, the gpt-5.6-sol model page.

Routing tip: keep GPT-class for product chat defaults; shift bulk classification, tagging, and draft rewrites to gpt-5.6-luna or deepseek-v4.1-flash-1017 on the same key. Escalate only failing slices to Terra, Sol, or gpt-6-astra after automated validators reject output. For Claude dollars see the Claude API pricing page; for Gemini see the Gemini API pricing page; for DeepSeek see the DeepSeek API pricing page; for Grok see the Grok API pricing page; for free prototyping see the gift-credit models page; for a side-by-side catalog see the model comparison page.

Cost hygiene that actually moves the bill: truncate unused chat history, prefer structured short outputs for batch jobs, cache identical prompts by hash, and measure tokens from the response usage object rather than guessing. Confirm live Keyo sell rates on Model Square after every catalog change.

GPT 6.1 Sol pricing and Sol vs Astra

gpt-6.1-sol lists $0.3677 input and $1.8385 output per million tokens in the default group checked on 2026-10-13. A 10,000-input and 2,000-output request is about $0.007354 before retries and extra charges. See the gpt-6.1-sol model page for the model guide and GPT-6.1 Sol vs GPT-6 Astra for the complete side-by-side table. Test accepted results on your own workload before deciding which tier to use.

# cURL — gpt-6-astra on Keyo
curl https://www.keyoapi.xyz/v1/chat/completions \
  -H "Authorization: Bearer $KEYO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-6-astra","messages":[{"role":"user","content":"Hello from GPT-6 Astra"}]}'

# Python — OpenAI SDK pointed at Keyo
from openai import OpenAI
client = OpenAI(base_url="https://www.keyoapi.xyz/v1", api_key="YOUR_KEYO_API_KEY")
r = client.chat.completions.create(
    model="gpt-6-astra",
    messages=[{"role": "user", "content": "Summarize this log line in one sentence."}],
)
print(r.choices[0].message.content)

Get started in five steps

1. Create an account — new users receive $10 of gift credit for prototyping.
2. Create an API key in the console.
3. Point your OpenAI SDK at https://www.keyoapi.xyz/v1 with the model ID from this page, or call the dedicated REST path from the API handbook.
4. Track spend in the console Wallet — $1 equals 500,000 quota units, and gift credit burns before recharged balance.
5. Set a per-key spending cap before wiring production traffic.

A verified first call is the acceptance test for the setup: the response should carry the payload shape documented in the handbook, the Wallet ledger should show a deduction in quota units (500,000 units = $1), and repeated calls should deduct linearly at the listed rate. Any mismatch usually means the model ID or endpoint is wrong — check both against the pricing list before debugging client code.

FAQ

Also see the model comparison page, the gift-credit models page, the full pricing list, and gpt-6-astra · gpt-5.6-luna · OpenAI API alternative.