AI startup Augment scraps 'unsustainable' pricing, users say new model is 10x worse
- Reference: 1760531661
- News link: https://www.theregister.co.uk/2025/10/15/augment_pricing_model/
- Source link:
The startup was launched in April 2024, co-founded by Igor Ostrovsky (ex-Microsoft software engineer) and Guy Gur-Ari (ex-Google AI research), and backed by venture capital including investment from Eric Schmidt (former Google CEO). Its main product is Augment Code, which provides AI-powered chat, Next Edit code suggestions, inline code completions, and agentic AI programming, which can create an app from scratch.
Popular features in Augment Code include a Memories feature that persists context across conversations, and a 200K context window, meaning the AI is better informed about the codebase it is asked to work on.
[1]
The price increase follows an earlier hike just six months ago. Originally, there was a free community plan for individuals, a $30/user-month plan for professionals, and a $60 plan for enterprise, all of which provided unlimited chats and completions. In early May, this was [2]replaced by "new, simpler pricing" based on the number of messages successfully processed. Free users got 50 messages, a $50.00 developer plan 600 messages, $100 professional 1,500 messages, and $250 max plan 4,500 messages.
[3]
[4]
A developer [5]complained at the time that it "now costs more than Cursor and Windsurf combined." The free plan then disappeared, replaced by an indie plan with 125 messages at $20/month.
That was just the start. A [6]new post from CEO Matt McClernan said "the user message model isn't sustainable for Augment Code as a business."
[7]
The problem is that the message abstraction does not reflect the actual AI usage, he explained, since a complex prompt could involve a lot of backend processing. He said one user on the $250 max plan is costing the company "approaching $15,000 per month," though it is not clear whether this user also purchased additional messages.
The latest model is based on credits, which are intended to reflect the actual cost of processing prompts. Since there is no exact mapping from messages to credits, the impact on users is variable, but McClernan noted that "our heaviest users will likely feel the change in price the most."
[8]KuzuDB says so long and thanks for all the commits, marooning community
[9]Bun 1.3 stuffs everything and kitchen sink into JS runtime
[10]Meta will move React to Linux Foundation to address vendor dominance fears
[11]Python releases version 3.14 – with cautious free-threaded support
Users have done their own calculations. One was [12]informed by email that in the previous seven days they had used "31 messages, corresponding to 40,982 credits under the new pricing model." This works out to a price increase of more than ten times. "I'm out. It was good while it lasted," they said.
Another [13]reaction was that Augment had exploited early users to refine the system and is now pricing them out. "We tested and optimized their infrastructure and paid for the privilege, and now are tossed aside."
McClernan argued that usage-based pricing is "fast becoming the industry standard," referencing pricing changes from competitors including Zed, Replit, Cursor, and Anthropic.
[14]
It appears that Augment entered this market with an unrealistic pricing model; two huge price rises within six months suggest a substantial miscalculation.
That said, it is also a wake-up call for users who do not appreciate the high cost of AI processing, which is compute-intensive.
A lesson here is the importance of AI cost optimization. Developers can craft prompts that use fewer tokens and tune models to reduce usage. Models also vary in cost from their providers. In an enterprise context, though, persuading developers to consider AI cost optimization when they are focused on getting their coding done may be challenging. ®
Get our [15]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aO_FFXKSyOPwH7CFouRmkAAAAVI&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://www.augmentcode.com/blog/new-simpler-pricing-with-user-messages
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aO_FFXKSyOPwH7CFouRmkAAAAVI&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aO_FFXKSyOPwH7CFouRmkAAAAVI&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://medium.com/vibe-coding/augmentcode-new-pricing-its-ridiculous-7de26c486115
[6] https://www.augmentcode.com/blog/augment-codes-pricing-is-changing
[7] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aO_FFXKSyOPwH7CFouRmkAAAAVI&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[8] https://www.theregister.com/2025/10/14/kuzudb_abandoned/
[9] https://www.theregister.com/2025/10/13/bun_13_full_of_features/
[10] https://www.theregister.com/2025/10/09/meta_react_foundation/
[11] https://www.theregister.com/2025/10/08/python_314_released_with_cautious/
[12] https://old.reddit.com/r/AugmentCodeAI/comments/1nzr1sc/is_augment_code_still_worth_it_after_the_price/ni48f35/
[13] https://old.reddit.com/r/AugmentCodeAI/comments/1nzr1sc/is_augment_code_still_worth_it_after_the_price/ni5b7rw/
[14] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aO_FFXKSyOPwH7CFouRmkAAAAVI&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[15] https://whitepapers.theregister.com/
Re: So
Sure it shows a benefit to the users... It clearly screams: Using one's own brain is much cheaper!
300x for starters
Muntz "Ha Ha" comes to mind. For people so smart they are incredibly naive, even stupid! Aside from worthless chat bits, you're not going to see prices below 1000/mo for any useful work. If I even stipulate they are capable of such. And the economic reality was bloody obvious months ago if you were paying the slightest attention.
If you base your calculations for your "business"
...on current or past AI pricing , you're a fucking idiot.
These companies lose billions of Dollars every week and you think they are just going to keep it the same?
Your first hit is free etc. etc.
Re: If you base your calculations for your "business"
Agree.
The companies probably did not intend to hurt their users, but this is the necessary effect of AI being _much_ more pricey to setup and run than their marketing tries to make you believe.
Every time you see a "$ XXX Billion investment in new AI infrastructure" headline, ask yourself, what part of that infrastructure (plus return on their "investment") you will be expect to pay for.
Either through taxes, because your government has been hooked by AI-services or by increased prices for products from companies who have fallen for AI - or, as in this case, directly as a consumer of AI coding tools.
35k ante
A decently useful rig costs that much. If you can get say 5 devs to use it cooperatively that's 200/mo/dev assuming no special power or Hvac needs and youre willing to ammortize for 3 loooooong years. I've built rigs for 5k that did a decent enough job for piddlyass usage.
Running your own infra or renting it straight rate $4/hr (1000/mo) is the only sane way forward. openAI et. Al. are terminally f'ked.
A b200 is 250k. At 2 yr depreciation, 50 active users it's $1/hr/user. Power and cooling about $1500/mo
Re: 35k ante
Out of morbid curiosity, what did these systems actually do? Yes, “AI”. But what was the application? And was it something that couldn’t be done via Google search or Stack Overflow search?
‘One user paying y is costing us x. Where y is greater than x’. I’d suggest that’s not the fault of the user, that’s a poor business plan, or technical implementation, or both.
AI is your lazy route to
market, code, documents, leads, sales... insert whatever you like.
It really doesn't matter, every bell end on the planet thinks they can invent an AI version of the field they work in to entice the other not so good bell ends to use the "AI" tool to do the job they're too thick to do with their own brain.....
This not the AI you're looking for.......
'He said one user on the $250 max plan is costing the company "approaching $15,000 per month," '
How many humans could you hire for $15,000 per month? They presumably would do a better job of it that this supposed A"I" thing.
Is there a snake in here?
Why do I hear hissing?
Has something sprung a leak?
End is nigh
I recon it starts to show material weakness by next summer. And completely implodes a year or so later. There is no saving this colossal mal-investment.
>> How many humans could you hire for $15,000 per month?
One and a bit decent ones. Two sort of averagely decent...
Second huge increase in six months...
And so it begins.
And so the cycle starts
Now, with this price increase, all the other offerings can start the process of raising their prices to get nearer to their true running costs.
At some point, the cost is going to reach $4,000+ pm per person for an all-you-can-use solution, but before then, the solution is going to really have to enhance a developer's performance.
Re: And so the cycle starts
$4000?! One of their users was costing them $15,000!
I suspect if everybody used it to the max, as they want us to, then that's closer to the truth.
So
Another AI example where there is no net benefit to the users. What a shock.