Learning Library

← Back to Library

Anthropic Launches Claude 3.5 Sonnet

Key Points

  • Anthropic released an upgraded Claude 3.5 Sonnet—keeping the same name but delivering substantially better performance, especially on coding evaluations.
  • They also launched a faster Haiku model that matches the quality of the older Opus version, indicating a shift toward consistent naming conventions.
  • Anthropic’s API business, though smaller in absolute terms than OpenAI’s, is growing about five‑times year‑over‑year and is on track to surpass OpenAI’s API revenue.
  • The new release emphasizes developer tooling, including a beta feature that lets Claude Sonnet control a computer screen for agentic workflows such as automated user‑acceptance testing in Replit’s “repet” agent.

Full Transcript

# Anthropic Launches Claude 3.5 Sonnet **Source:** [https://www.youtube.com/watch?v=GMyJSd2yJYA](https://www.youtube.com/watch?v=GMyJSd2yJYA) **Duration:** 00:03:00 ## Summary - Anthropic released an upgraded Claude 3.5 Sonnet—keeping the same name but delivering substantially better performance, especially on coding evaluations. - They also launched a faster Haiku model that matches the quality of the older Opus version, indicating a shift toward consistent naming conventions. - Anthropic’s API business, though smaller in absolute terms than OpenAI’s, is growing about five‑times year‑over‑year and is on track to surpass OpenAI’s API revenue. - The new release emphasizes developer tooling, including a beta feature that lets Claude Sonnet control a computer screen for agentic workflows such as automated user‑acceptance testing in Replit’s “repet” agent. ## Sections - [00:00:00](https://www.youtube.com/watch?v=GMyJSd2yJYA&t=0s) **Anthropic's Strategic Claude 3.5 Release** - Anthropic unveiled an upgraded Claude 3.5 Sonnet—keeping the familiar name but delivering markedly better evaluation scores and coding ability—to signal a market‑share play against OpenAI while emphasizing its rapidly growing API business. ## Full Transcript
0:00anthropic dropped a huge release I want 0:02to get into it and I want to tell you 0:03kind of how it shapes our understanding 0:05of the players in this field because 0:06anthropic and open AI are starting to 0:08take up different niches and we need to 0:10understand that so first what did they 0:12drop they dropped an upgraded Claude 3.5 0:16Sonet they did not rename it they did 0:18not give it a new number again these 0:19naming conventions are bizarre it's 3.5 0:22sonnet just like you've always known and 0:24loved except it's a lot better like it's 0:26substantially better on the evals it's 0:28way better at coding which was always 0:30good at and that's going to come back 0:32around you'll see why in a second and so 0:35they dropped this new thing they give it 0:36the same name for reasons I don't fully 0:38understand they then drop an upgraded 0:40version of haiku which they say is as 0:43good as their old version of Opus but 0:45much much faster we are starting to see 0:48that anthropic has picked naming 0:50conventions and are going to stick to 0:52them and so I 0:55think this may be a bit of a signal to 0:58the other major players in the field to 0:59gole gole and to open AI that anthropic 1:02has more in the tank that they are not 1:04releasing a new named model deliberately 1:06and that this should be viewed as a 1:08waiting game release to continue to gain 1:11market share where they play best and 1:13that gets to the strategy here so 1:16anthropic is about a quarter of the size 1:18of open AI but their API business is 1:20much stronger their API business is 1:23growing much faster than open ai's API 1:26business and is on Pace to overtake it 1:28so year-over-year open AI uh open AI is 1:31doubling its API business and they're 1:33over a billion dollars well anthropic is 1:35at somewhere around a $700 million run 1:38rate but they're growing 5x 1:39year-over-year like they're growing much 1:40much faster on that API side and that is 1:42why in their press release they 1:45highlighted the dev tooling capacities 1:47they made Sonet better for coders they 1:49made sure that they called out how Dev 1:53tooling is affected by their new 1:55experimental computer screen usage and 1:58this is this is the big thing right they 2:01are going to let Claude Sonet control 2:03your computer screen it's the first hint 2:05we've seen of where these agentic 2:07workflows are going to go it's still in 2:09public beta it's not fully released yet 2:12but people can access it and use it and 2:14the use case they call out is developer 2:16friendly they are saying that repet is 2:18using this in repet agent which is their 2:21AI Builder to help user acceptance 2:24testing for things that users are 2:26building so that the repet agent 2:27actually goes and like tests it on a 2:28screen the way you would 2:30um and improves the quality of the code 2:32that way so they're doubling down on 2:35their API positioning I think that's 2:37smart IT differentiates them from open 2:39AI they are not changing the name for 2:41reasons I don't fully understand but it 2:44is a big release uh it's substantially 2:46better on the coding evals we'll have to 2:49see how it looks in the wild uh but I'm 2:51excited to play with Sonet 3.5 the new 2:53version uh new and improved right so let 2:56me know what you're doing let me know 2:57how you're coding cheers