Deepseek Says Its Hit AI Model Cost Just $294k to Train

Posted4 months agoActive4 months ago

jonbaer

19 points

9 comments

reuters.comTechstory

skepticalmixed

Debate

60/100

AI Training CostsLLM DevelopmentGPU Costs

Key topics

AI Training Costs

LLM Development

GPU Costs

DeepSeek claims to have trained its AI model for $294k, sparking discussion about the validity of this claim and its implications for the AI market, particularly regarding GPU costs.

Snapshot generated from the HN discussion

Discussion Activity

Light discussion

First comment

39m

Peak period

2-3h

Avg / period

1.1

Key moments

01Story posted
Sep 19, 2025 at 12:24 AM EDT
4 months ago
Step 01
02First comment
Sep 19, 2025 at 1:03 AM EDT
39m after posting
Step 02
03Peak activity
2 comments in 2-3h
Hottest window of the conversation
Step 03
04Latest activity
Sep 19, 2025 at 1:32 PM EDT
4 months ago
Step 04

Generating AI Summary...

Analyzing up to 500 comments to identify key contributors and discussion patterns

Discussion (9 comments)

Showing 9 comments

onion2k

4 months ago

2 replies

Maybe, if you don't include the >$10m investment in H800 hardware. Still a lot cheaper than competitors though.

48terry

4 months ago

1 reply

Yes, if we include a cost they didn't include, the cost would be different.

beaner_count

4 months ago

More like, if you exclude costs, things cost whatever you want to tell people they cost.

jml7c5

4 months ago

1 reply

No, their calculation is based on a rental price of $2/hour.

yorwba

4 months ago

1 reply

Right, but they didn't use rented GPUs, so it's a purely notional figure. It's an appropriate value for comparison to other single training runs (e.g. it tells you that turning DeepSeek-V3 into DeepSeek-R1 cost much less than training DeepSeek-V3 from scratch) but not for the entire budget of a company training LLMs.

DeepSeek spent a large amount upfront to build a cluster that they can run lots of small experiments on over the course of several years. If you only focus on the successful ones, it looks like their costs are much lower than they were end-to-end.

yunohn

4 months ago

1 reply

No, they’re saying training a model, specifically DeepSeek, costs X using N hrs of Y GPU rental.

yorwba

4 months ago

If by "they" you mean DeepSeek, they're not saying this, since you might not actually be able to rent a cluster of 512 H800s wired together with high-bandwidth interconnects at that GPU-hour price point. If you rent smaller groups of GPUs piecemeal in different locations and try to transfer weight updates between them over the internet, it'll kill your throughput.

PaulRobinson

4 months ago

1 reply

They made this claim in a peer reviewed paper submitted to Nature, but it’s not clear how peers could evaluate the truth of this claim.

If it’s true, and the consensus is that we are hitting limits of how to improve these models, the hypothesis that the entire market is in a bubble over-indexed on GPU costs [0] starts to look more credible.

At the very least, OpenAI and Anthropic look ridiculously inefficient. Mind you, given the numbers on the Oracle deal don’t add up, this is all starting to sound insane already.

[0] https://www.wheresyoured.at/the-haters-gui/

fspeech

4 months ago

These numbers were easily supported by those who attempted to replicate the RL portion of their work. The foundational model training is harder to verify but is also not central to the paper.

View full discussion on Hacker News

ID: 45297935Type: storyLast synced: 11/20/2025, 5:48:27 PM

Want the full context?

Jump to the original sources

Read the primary article or dive into the live Hacker News thread when you're ready.

Open link View on HN