this post was submitted on 06 Aug 2026
92 points (97.9% liked)

technology

24450 readers
349 users here now

On the road to fully automated luxury gay space communism.

Spreading Linux propaganda since 2020

Rules:

founded 6 years ago
MODERATORS
 

Dear DeepSeek API user,

We plan to raise the overall pricing for DeepSeek API services in the near future, with a significant increase expected. Please plan your usage accordingly. The specific pricing plan will be subject to official notice. Please keep an eye on the Open Platform announcements and check your email for further details. If you continue to use our services after the billing adjustment, you will be deemed to have accepted the adjusted billing terms. If you do not agree, you may choose to cancel your service and apply for a refund. Should you have any questions or require further information, please do not hesitate to contact us.

Thank you for your support and understanding! DeepSeek Team

This is from an email I got.

Couldn't really find reliable information about the price increase because everything about it is paywalled and the Deepseek website doesn't have info on the price increase.

all 33 comments
sorted by: hot top controversial new old
[–] RNAi@hexbear.net 40 points 2 weeks ago (2 children)
[–] jack@hexbear.net 27 points 2 weeks ago (1 children)

if even China can't make LLMs make economic sense, Zitron will ascend to a higer plane

[–] hello_hello@hexbear.net 15 points 2 weeks ago

We take good care to heed Ed Zitron-ist thought.

[–] ggtdbz@lemmy.dbzer0.com 13 points 2 weeks ago

Every time I listen to an interesting podcast from him that I think of sharing with people I work with he makes sure to immediately drop twelve expletives that I just do not want to share with these people.

That said, I’ve adopted “doesn’t know shit about fuck” for tactical usage and it’s beautiful

[–] DasRav@hexbear.net 39 points 2 weeks ago (1 children)

Deepseek is like 200 million in the red. Unlike other AI companies, they might be able to dig themselves out of that with a still affordable price hike. The average US company would need to 10x prices to do that.

[–] Soot@hexbear.net 40 points 2 weeks ago (1 children)

Interesting. 200 million is peanuts compared to most, OpenAI lost 9 billion last year.

[–] DasRav@hexbear.net 30 points 2 weeks ago

Those 200 million are an estimate of their total operational losses since 2023 as well. It's probably not the most accurate number, but absolutely close enough to illustrate just how totally, utterly wasteful the US AI corporations are being.

[–] CliffordBigRedDog@hexbear.net 37 points 2 weeks ago (1 children)

how about they deepseek muh ballz

[–] ClathrateG@hexbear.net 18 points 2 weeks ago* (last edited 2 weeks ago)

I tried a couple times for it to get a funny reply to this prompt but the most I got was 'it appears you are trying to make a joke involving a pun on my name'

[–] tamagotchicowboy@hexbear.net 35 points 2 weeks ago

They must be getting hammered, this was after they announced peak hours pricing.

[–] JoeByeThen@hexbear.net 30 points 2 weeks ago
[–] supafuzz@hexbear.net 28 points 2 weeks ago (1 children)

they've been practically giving it away, the v4 pro prices on openrouter are stupid cheap compared to kimi or glm or qwen. wasn't going to stay like that forever

[–] chgxvjh@hexbear.net 26 points 2 weeks ago

They pretty recently announced that the 50% off they had before will be the new normal price and now they raise the price. Their pricing strategy doesn't seem entirely thought through.

[–] Dort_Owl@hexbear.net 19 points 2 weeks ago
[–] Juice@midwest.social 15 points 2 weeks ago (1 children)
[–] chgxvjh@hexbear.net 14 points 2 weeks ago

I'd would like to know that too sadness

[–] LadyCajAsca@hexbear.net 15 points 2 weeks ago

I mean I still have other chinese AIs in mind (Kimi, Qwen, etc.) so it's not all bad but I wonder what the upcoming release of a new model has to do with this?

[–] insurgentrat@hexbear.net 15 points 2 weeks ago (1 children)

A tech company peddling dependence increased prices after being established? Woah

[–] Letztertod@hexbear.net 4 points 2 weeks ago* (last edited 2 weeks ago) (1 children)

People genuinely thought it being a Chinese company and China being a socialist country, it would be raining milk and honey from the sky for them.

[–] insurgentrat@hexbear.net 5 points 2 weeks ago (1 children)

People that think that hideously unethical slop machines but Chinese so lea-happy are fucking worse than techbros who are at least honestly garbage.

[–] Letztertod@hexbear.net 3 points 2 weeks ago

I genuinely agree, a tech bro is easy to spot and ignore but these infection every corner of your life as a leftist.

[–] Krem@hexbear.net 14 points 2 weeks ago

people are paying for ai?

[–] KssioAug@lemmy.dbzer0.com 11 points 2 weeks ago* (last edited 2 weeks ago) (2 children)

I hope they don't fuck up. Expecting a price increase from time to time is normal, but the wording (significant increase expected) is worrisome. It feels particularly bad after they just recently announced they'd keep the promotional price, and followed by increasing price in peak hours.

[–] BeanisBrain@hexbear.net 10 points 2 weeks ago

Best case scenario: they're deliberately underpromising and overdelivering, and the price increase is overstated.

[–] TraschcanOfIdeology@hexbear.net 5 points 2 weeks ago

They're not immune from the fundamental problem of the AI bubble: if people had to pay the true cost of the thing (even if Chinese LLMs are significantly more efficient), almost no use case would have a justifiable ROI. LLMs are expensive to train and run, and users have been shielded from that cost by VCs and CAPEX spending.

OpenAI and Anthropic did the same thing months ago. It is coming for all multi-purpose LLMs

[–] insurgentrat@hexbear.net 9 points 2 weeks ago (1 children)

If you're affected by this I hate you and I'm glad your slop addiction hurts

[–] freshfusion@discuss.tchncs.de 7 points 2 weeks ago (1 children)

Just use it offline bro if you can. I do not really use LLMs much lately so IDK, but I prefer gemma4 though. Deepseek distilled to 7 or 8b will probably work fine with mid range relatively recent hardware.

[–] Moidialectica@hexbear.net 4 points 2 weeks ago (2 children)

hows gemma 4 treating you lately? the prices on openrouter were worse than deepseek at the time so i never even bothered to try it

[–] pongo1231@hexbear.net 4 points 2 weeks ago

Quite efficient to run locally, 26b A4B runs pretty well on a 8GB card at around 10 t/s for me. Excells at tasks like translation but falls short at agentic coding in comparison to Qwen.

[–] alexandra_kollontai@hexbear.net 4 points 2 weeks ago

Gemma 4 26B A4B: Vision is very imprecise. Reasons forever. If it ever stops reasoning then the answers are usually good for a local model, but it also has surprisingly bad failure cases too. The usual. It's very fast.