GPT-6.2 Pricing: No Official Rates; Where Numbers Come From
As of 2026-11-02, the OpenAI API pricing page, the model catalog and the changelog carry no rate for GPT-6.2. The two sets of figures circulating online come from unconfirmed leaks, each with its site, account and date; the live GPT-6.x official table is below.
Updated 2026-11-02
Prices change: on QCode, Claude, GPT and Chinese models are billed per token — live rates on /models.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Four key points
GPT-6.2 official rates
The pricing page, the model catalog, the changelog and the live model pages carry no 6.2; checked 2026-11-02.
The gpt-6.2-astra model page
Its official model-page URL returns Not found, and so does the plain gpt-6.2 page.
The live 6.1 Sol rate
Short context, per 1M tokens: $2.00 in, $10.00 out.
The long-context threshold
Official rule: prompts with more than 272K input tokens are priced at 2x input and cache rates and 1.5x output for the full request.
What can actually be checked about GPT-6.2 pricing
As of 2026-11-02 exactly one thing can be verified word for word: OpenAI has published no unit price for GPT-6.2. The pricing page splits its tables into Standard, Batch, Flex, Fast and Ultrafast, and those tables list gpt-6-astra, gpt-6-sol, gpt-6.1-sol, gpt-6.1-terra, gpt-6-luna and the previous GPT-5.6 generation — no row for 6.2. The newest general model in the catalog is GPT-6.1 Terra, released 2026-10-26; the last model entry in the changelog is GPT-6 Rosalind of 2026-10-30, which OpenAI describes as built for life-sciences research and gated behind the trusted-access program.
The last two official model entries: Terra (2026-10-26), Rosalind (2026-10-30)
The official changelog records GPT-6.1 Terra on 2026-10-26, positioned for visual reasoning, computer use and natural writing, and GPT-6 Rosalind on 2026-10-30. Neither entry names 6.2 or gives it a rate, and the pricing page and model catalog say the same.
Price timeline (dates and figures from the official changelog)
Official changelog: GPT-5.6 Sol costs $4 input and $20 output per 1M tokens, with the promotional pricing available at least through November 21, 2026.
OpenAI releases GPT-6 Sol and GPT-6 Luna; the changelog lists Sol at $2 input, $0.20 cached input and $10 output, and the $0.10 / $0.01 / $0.50 row as Luna.
GPT-6.1 Sol ships; the changelog spells out standard pricing per 1M tokens: $2 input, $0.10 cached input, $2.50 cache write, $10 output.
What OpenAI has published, and what it has not
Confirmed (verbatim on official pages, 2026-11-02)
Standard short-context rates on the OpenAI pricing page (up to 272K input tokens, per 1M tokens; cache writes at the 30-minute tier): gpt-6-astra $10.00 input, $1.00 cached input, $12.50 cache write, $50.00 output; gpt-6-sol $2.00 / $0.20 / $2.50 / $10.00; gpt-6.1-sol $2.00 / $0.10 / $2.50 / $10.00; gpt-6.1-terra $1.00 / $0.05 / $1.25 / $5.00; the $0.10 / $0.01 / $0.125 / $0.50 row is gpt-6-luna. The model pages add: prompts with more than 272K input tokens are priced at 2x input and cache rates and 1.5x output for the full request; Batch and Flex are priced at 50% of Standard rates, and Fast mode is priced at 2x the applicable rates.
Unconfirmed: the two pairs of figures in circulation
The first pair comes from a kie.ai article dated 2026-11-01 by Sofia Marenco: it calls $4.00 input / $20.00 output “leaked data”, labels the same figure an “Unconfirmed leak (@SNARKAMOTO)”, and states that OpenAI had not released GPT-6.2 or listed it in the public API catalog as of 2026-11-01. OpenAI has not confirmed it. The second pair, $20 input / $100 output, appears in the Bel / Betelgeuse claim table; the AI IDE List fact-check (published 2026-10-31, updated 2026-11-01) keeps it unconfirmed and writes: “The alleged $20/$100 pair also matches GPT-6 Astra’s published Fast-mode rates.” On the official side, the model page says “Fast mode is priced at 2x the applicable rates.”, and the Fast table on the pricing page shows gpt-6-astra at $20.00 input and $100.00 output. The same fact-check also writes that this overlap neither authenticates nor disproves the rumor, and OpenAI has not confirmed the figures. Separately from pricing, @AGTPinsights wrote on 2026-11-01 that gpt-6.2-astra had appeared in Codex usage analytics, and the kie.ai article says developers noticed it on 2026-10-31 and 2026-11-01 through usage accounting logs. OpenAI has not confirmed it, and no catalog lists that id.
Published rates versus numbers that are not a rate card
Published by OpenAI (per 1M tokens, short context)
The GPT-6 Astra model page lists $10 input and $50 output at standard, and says Fast is priced at 2x the applicable rates; the Fast table on the pricing page shows exactly $20.00 in and $100.00 out for this row.
The two unconfirmed pairs
The $4 in / $20 out reported by kie.ai (2026-11-01, Sofia Marenco) and the $20 / $100 in the Bel claim table are unconfirmed by OpenAI, and no row labelled 6.2 exists on the official pricing page.
What to check the day OpenAI publishes a GPT-6.2 rate card
①The four standard items: input, cached input, cache write and output — the official pricing page prices each separately, and cache writes split into 30-minute and 12-hour retention. ②The long-context column: the page labels up to 272K input tokens “Short context” and everything above “Long context”, with the whole request re-priced. ③Processing tiers that exist in today's official table: Batch, Flex, Fast and Ultrafast — the 6.1 Sol model page states Ultrafast at 6x Standard, Fast at 2x Standard and Batch and Flex at 50% of Standard. ④Whether regional processing adds a 10% premium. ⑤Whether there is a distinct model id and snapshot name, and whether the model lands in the API, in Codex or in Chat. ⑥On the QCode side: per-model rates and tiers follow the /models page.
Reading prices on QCode
QCode bills per token, with per-model rates on /models. New-model availability follows /models. For a live GPT-6.x today, /models lists gpt-6-astra, gpt-6-sol, gpt-6.1-sol and gpt-6.1-terra; QCode does not currently support gpt-6-luna, so please use another GPT model. Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; if you already have a Personal Plan, add a booster pack when you need more quota.
Questions
Does GPT-6.2 have an official API price yet?
No. As of 2026-11-02 the OpenAI pricing page, model catalog and changelog show no rate for 6.2, and the gpt-6.2-astra model page returns 404.
Where do the circulating $4 input and $20 output figures come from?
From a kie.ai article dated 2026-11-01 by Sofia Marenco, which itself labels them an unconfirmed leak attributed to @SNARKAMOTO; OpenAI has not confirmed them.
And the other pair, $20 / $100?
It sits in the Bel / Betelgeuse claim table; the AI IDE List fact-check (published 2026-10-31, updated 2026-11-01) keeps it unconfirmed and points out that it overlaps with the published Fast-tier rate of GPT-6 Astra.
Can I budget GPT-6.2 at the Astra Fast rate?
No. None of the OpenAI official pages checked puts 6.2 next to a price in any sentence, and the AI IDE List fact-check (updated 2026-11-01) states that the overlap neither authenticates nor disproves the claim. Only released models can be scheduled today.
Will this page be updated?
Once OpenAI publishes a 6.2 rate card, this page is refreshed against the official pricing and model pages, including the check date; until then the verifiable status is: unpublished.
Can I call GPT-6.2 through QCode today?
As of 2026-11-02, QCode's /models carries no GPT-6.2; models and per-token rates follow that page.
Sources
OpenAI developer docs: the API pricing page, the model catalog, the model pages for GPT-6 Astra, GPT-6 Sol, GPT-6.1 Sol, GPT-6.1 Terra and GPT-6 Luna, and the API changelog (all captured 2026-11-02); the kie.ai article of 2026-11-01; the AI IDE List fact-check published 2026-10-31 and updated 2026-11-01; an @AGTPinsights post on X dated 2026-11-01; qcode.cc /models and the Pay-as-you-go Packs page (captured 2026-11-02).
Live GPT-6.x rates, billed per token
Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; if you already have a Personal Plan, add a booster pack when you need more quota.
Related reading
GPT-6.2 Release Date: Official Status So Far
Where 6.2 stands and which date claims come from unconfirmed posts.
GPT-6 Sol / Luna price tracker
Sol at $2 / $10, the Luna tier, plus the official cache, long-context, Batch and Fast rules.
QCode pricing: Personal Plans and Pay-as-you-go Packs
Personal Plans give a daily quota for 30 days; a Pay-as-you-go Pack is a one-time quota you use within 7 days.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
OpenAI figures and dates on this page were checked on 2026-11-02 against the official pricing page, model pages and changelog; official pages remain authoritative. Numbers marked unconfirmed come from third-party articles and are not a rate card. QCode models and per-token rates follow /models; official changes are not announced separately.