OpenAI Flex pricing

[miloszkukla]

🏷️ Discussion Type

Question

Body

Do you plan to add a cheaper GPT 5.5 per usage pricing option (that trades-off latency) that would correspond to https://developers.openai.com/api/docs/pricing?latest-pricing=flex ?

[Mister-Halder]

Yes — this already exists via OpenAI’s Flex processing tier.
It offers lower per-usage cost (≈50% cheaper) in exchange for higher latency and less guaranteed availability, which matches the “cheaper but slower” model you’re asking about.

[miloszkukla]

I'm asking about Copilot pricing, not OpenAI pricing to avoid BYOK

[mecodeatlas]

Hi @henock-lobu, thanks for participating!

Unfortunately, we don’t currently have moderators for languages other than English. Until that changes, we need to ask that everyone use English here in the GitHub Community when posting to ensure discussions are adhering to our Code of Conduct.

We’ll be hiding this comment for now, but please feel encouraged to create a new one in English :)

[Janviagrawal4127]

something very close to what you’re asking already exists—but it’s not a separate “cheaper model,” it’s a pricing mode (Flex/Batch).

OpenAI does offer lower-cost tiers (Flex / Batch) that trade off latency and availability for price.
These modes are typically ~50% of standard pricing for models like GPT-5.5.
For example, GPT-5.5 standard is about $5 input / $30 output per 1M tokens, while Flex/Batch brings it down to roughly $2.5 / $15.

What Flex actually means:

Slower responses
Possible queueing or delayed execution
Best suited for non-real-time workloads (batch jobs, evaluations, data processing)
So to your question:
✅ A cheaper, latency-tradeoff option already exists (Flex)
❌ There’s no separate “GPT-5.5 cheap version” beyond pricing tiers announced right now

If you’re expecting something like a distinct lower-tier API plan (e.g., per-request discounted real-time usage) — that hasn’t been officially announced. The current strategy is:

same model → different pricing tiers (standard, flex, batch, priority)

[Janviagrawal4127]

there’s no separate announcement of a new cheaper GPT-5.5 tier beyond what’s already available.
OpenAI’s current approach is to offer pricing modes (like Flex/Batch) rather than separate “cheaper versions” of the same model. These options already provide lower per-usage cost in exchange for higher latency or delayed processing, and are intended for non-real-time workloads.
So:

✅ Lower-cost, latency trade-off options do exist (e.g., Flex/Batch)

❌ No publicly announced plan for a new distinct cheaper GPT-5.5 pricing tier beyond those modes

For the most accurate and up-to-date information, it’s best to keep an eye on official pricing updates in the OpenAI API documentation.

[miloszkukla]

@mecodeatlas could you please recommend who from the GitHub company to tag to get the answer? (sadly existing answers are totally off-topic, AI generated probably)