DeepSeek, China's AI model: News & Discussion
- Thread starter Beijingwalker
- Start date
冖_冖
Registered Member
OpenAI and Anthropic face new AI reality as users shift from 'tokenmaxxing' to efficiency
Companies are tightening their AI budgets to focus on getting a return on their investment, and that could dampen growth rates at OpenAI and Anthropic.
www.cnbc.com
Flo Crivello’s expenses were out of whack, and there was only one way to get them under control.
Earlier this month, the CEO of AI startup Lindy switched his company off Anthropic’s Claude models, moving 100% of its traffic to DeepSeek, a Chinese company that makes cheaper, open-weight alternatives.
“We did it, and you could see that cost curve go down, like, crash to the ground,” Crivello said in an interview from his company’s San Francisco headquarters. He said the decision will save Lindy millions of dollars within months, though he still expects the roughly 25-person company to spend more on AI than payroll.
“It’s a matter of survival for the business,” Crivello said. “That’s all it is.”
lightning f57
Trusted Member
Alibaba released some very good open source models which were distilled down to smaller models that can fit on your PC.
Unfortunately the team which made some brilliant small models has lost some key members.
I hope they can continue to release open source models along with other companies.
Unfortunately the team which made some brilliant small models has lost some key members.
I hope they can continue to release open source models along with other companies.
冖_冖
Registered Member
Coinbase Switches to Cheaper Chinese AI Models, Halving Its AI Bill - Lapaas Voice
Coinbase now runs on cheaper Chinese AI models like GLM 5.2 and Kimi 2.7, halving its AI bill while usage climbs, piling pricing pressure on Western AI labs.
voice.lapaas.com
U.S. tech firms quietly pivot to Chinese AI models, with Coinbase leading the adoption of GLM and Kimi
Coinbase Switches to Cheaper Chinese AI Models, Halving Its AI Bill
Coinbase is a big US company. People use it to buy and sell cryptocurrency, which is a kind of digital money. Now Coinbase runs on cheap AI made in China. And this change cut its AI bill in half. The boss of Coinbase is Brian Armstrong. (A boss who runs a whole company is called the CEO, short for chief executive officer.) He says the company uses more AI than ever, but pays much less.
Coinbase now uses AI models called GLM 5.2 and Kimi 2.7. Both were made by labs in China. Armstrong said the company now uses more “tokens” than ever. But it pays about half of what it used to pay. A token is a tiny piece of text. It is about one word, or part of a word. AI companies charge you based on how many tokens you use. So more tokens normally means a bigger bill. But Coinbase did the opposite. It used more AI and still paid less.
Coinbase workers can still pick any AI model they like. But 91% of them never even reach their old usage limits, Armstrong said. So for most jobs, the cheaper Chinese models work just fine.
神威98
Registered Member
It was more like Deepseek banning discussions on "trade secrets". Last year, I asked Deepseek what chart settings were used by institutional traders and hedge funds (there must be a standard to what they are using as I don't think it be a good idea if the chart looked different from one trader to the next). It went through all the employee manuals of all major funds around the world and told me what those standards were down to their exact values. And it was the same whether the fund was located in NY or Tokyo. They were all using the same friggin settings.
This year, I asked question on this topic from another angle. Deepseek replies no info or does not know what I'm talking about. I think they started blacklisting this kind of "insider" stuff that came from manuals written for employees and analysts and professional traders of hedge funds from public knowledge...
ChillVein
Registered Member
China's AI Poses Massive Pricing Challenge for Silicon Valley. One of the world's largest cryptocurrency companies, Coinbase, has shifted its routine AI workloads from costly U.S. frontier models to open-source models from Zhipu and DeepSeek. According to CEO Brian Armstrong, this shift has slashed internal AI costs by nearly 50%.Reports indicate that running the same enterprise workloads on Anthropic's Claude costs about $4,811, while OpenAI's GPT-5.5 runs at $3,357.DeepSeek V4 costs around $1,071, and Zhipu's GLM 5.2 drops the price to just $544.
Users who are viewing this thread
Pakistan Defence Latest
-
-
Bangladesh Army’s new battalion has Umar, Abu Bakr, Ali and Usman companies (3 Viewers)- Latest: Areesh
-
-
-
Country Watch Latest
Latest Posts
-
-
-
-
-
Air India to cut international flights till July amid fuel surge, airspace curbs (21 Viewers)- Latest: Master Chief