Impact of OpenAI's 80% Price Drop on Lindy Credit Usage

Ben S. · 2025-06-11 08:00:16.05782+00

Hey all! I just saw that OpenAI has dropped the price of o3 by 80% and so, I was wondering: How does Lindy handle changes like these in regards to credit usage? I.e. will the amount of credits we use during an o3 query now drop dramatically? Thanks! Cc

Impact of OpenAI's 80% Price Drop on Lindy Credit Usage | Lindy Community

15 comments

· Sorted by Oldest

Marvin A.
·
·
Yes. 🤑 🤑 🤑 announcement coming soon
🙌1
Ben S.
·
·
Awesome! Thanks for coming back on this Marvin A.! So, is this a general thing that happens by default i.e. any time any model gets cheaper, the team update? Or is it something we need to give a nudge on? As a slightly separate note, I was surprised to notice when I was playing with the models the other day that GPT 4o mini (the tiniest model) was costing the same as claude 3.7 - both 1 credit! Would be interesting to hear your thoughts on this one - is 1 credit the smallest denomination?
👀1
Marvin A.
·
·
yeah usually no nudge needed — it’s not a change by default but we keep an close eye on model prices and make sure to adjust accordingly 🙂
🙌1
Marvin A.
·
·
interesting re claude 3.7 — can you share the task URL? would love to take a look
Marvin A.
·
·
How to share your Lindy task URL: 1. Click your Lindy agent → click “tasks” 2. Find the task you need help with 3. Click the specific task (you should see the full execution) 4. Copy that page URL from your browser 5. Paste the URL in your support request
Ben S.
·
·
Awesome, thanks Marvin A.! I sent you a DM with the links. Hope that works okay, keen to hear your thoughts!
✅1
Ben S.
·
·
Basically, the underlying question here - is 1 credit the bare minimum, or can we use super basic models at even smaller prices (otherwise every action effectively costs $0.01, which isn't necessarily scalable)
Marvin A.
·
·
Just replied via DM but the tl;dr: yes most actions consume a minimum of 1 credit. There are some outliers though (e.g. Perplexity starting at 0.5 credits, some of the LinkedIn scrapers start at 0.1 credits)
Loading attachments
🙏1
Ben S.
·
·
Thank you!! User feedback from me would be that I'd love to have a 'super cheap' option - one that can do very basic stuff at a very low cost. Happy to give examples if useful!
🔥1
Marvin A.
·
·
yes please! can you share a couple. makes it much more tangible for me to share with the product team 🙂 🙏🏾
Derek S.
·
·
Hey when can we use our own API keys? The credit costs are frustrating - I'm seeing o3 spending a ton of credits today.
Marvin A.
·
·
hey derek, wait what means “a ton of credits”? we literally cut o3 costs by 80% 😄
Marvin A.
·
·
happy to take a look if you can share a task URL! 👀
Marvin A.
·
·
Re: bring your own API keys - we’re definitely planning to add this feature, but it’s not in the immediate roadmap. I don’t want to commit to a specific timeline since these things can shift, but it’s on our list! 🙏🏾
Ben S.
·
·
Hey Marvin A., as promised, sharing a couple of use case RE our chat above below, I'll add more if they come to me. Quick caveat - maybe some of these can be handled with existing built in functionality and it's just my inexperience, feel free to highlight if so!
1.
Basic text tidy up - let's say I want to parse a users dumb text dump into something more legible/tidy, I wouldn't want to use a big/expensive model for that.
2.
Extracting - i.e. if I want to extract a specific word/list item from a paragraph.
3.
Basic mail merge - if I want to simply dynamically insert the name of the person I'm sending an email to (this could actually be really powerful because then I could use Lindy for all email functionality instead of having something separate). At scale, even a small price per credit would make this prohibitive.
I'm sure I can come up with more (and will add more if they come to me, just super busy right now), but generally speaking, the benefit of having a 'dumb' model that's super cheap (maybe like Groq or something open source?) which is included unlimited in the paid subscription would be SO good and would embed my business even deeper into Lindy (thus increasing retention).

Marvin A.
·
·
Yes. 🤑 🤑 🤑 announcement coming soon
🙌1
Ben S.
·
·
Awesome! Thanks for coming back on this Marvin A.! So, is this a general thing that happens by default i.e. any time any model gets cheaper, the team update? Or is it something we need to give a nudge on? As a slightly separate note, I was surprised to notice when I was playing with the models the other day that GPT 4o mini (the tiniest model) was costing the same as claude 3.7 - both 1 credit! Would be interesting to hear your thoughts on this one - is 1 credit the smallest denomination?
👀1
Marvin A.
·
·
yeah usually no nudge needed — it’s not a change by default but we keep an close eye on model prices and make sure to adjust accordingly 🙂
🙌1
Marvin A.
·
·
interesting re claude 3.7 — can you share the task URL? would love to take a look
Marvin A.
·
·
How to share your Lindy task URL: 1. Click your Lindy agent → click “tasks” 2. Find the task you need help with 3. Click the specific task (you should see the full execution) 4. Copy that page URL from your browser 5. Paste the URL in your support request
Ben S.
·
·
Awesome, thanks Marvin A.! I sent you a DM with the links. Hope that works okay, keen to hear your thoughts!
✅1
Ben S.
·
·
Basically, the underlying question here - is 1 credit the bare minimum, or can we use super basic models at even smaller prices (otherwise every action effectively costs $0.01, which isn't necessarily scalable)
Marvin A.
·
·
Just replied via DM but the tl;dr: yes most actions consume a minimum of 1 credit. There are some outliers though (e.g. Perplexity starting at 0.5 credits, some of the LinkedIn scrapers start at 0.1 credits)
Loading attachments
🙏1
Ben S.
·
·
Thank you!! User feedback from me would be that I'd love to have a 'super cheap' option - one that can do very basic stuff at a very low cost. Happy to give examples if useful!
🔥1
Marvin A.
·
·
yes please! can you share a couple. makes it much more tangible for me to share with the product team 🙂 🙏🏾
Derek S.
·
·
Hey when can we use our own API keys? The credit costs are frustrating - I'm seeing o3 spending a ton of credits today.
Marvin A.
·
·
hey derek, wait what means “a ton of credits”? we literally cut o3 costs by 80% 😄
Marvin A.
·
·
happy to take a look if you can share a task URL! 👀
Marvin A.
·
·
Re: bring your own API keys - we’re definitely planning to add this feature, but it’s not in the immediate roadmap. I don’t want to commit to a specific timeline since these things can shift, but it’s on our list! 🙏🏾
Ben S.
·
·
Hey Marvin A., as promised, sharing a couple of use case RE our chat above below, I'll add more if they come to me. Quick caveat - maybe some of these can be handled with existing built in functionality and it's just my inexperience, feel free to highlight if so!
1.
Basic text tidy up - let's say I want to parse a users dumb text dump into something more legible/tidy, I wouldn't want to use a big/expensive model for that.
2.
Extracting - i.e. if I want to extract a specific word/list item from a paragraph.
3.
Basic mail merge - if I want to simply dynamically insert the name of the person I'm sending an email to (this could actually be really powerful because then I could use Lindy for all email functionality instead of having something separate). At scale, even a small price per credit would make this prohibitive.
I'm sure I can come up with more (and will add more if they come to me, just super busy right now), but generally speaking, the benefit of having a 'dumb' model that's super cheap (maybe like Groq or something open source?) which is included unlimited in the paid subscription would be SO good and would embed my business even deeper into Lindy (thus increasing retention).