tweetindex
EN

Jun Song ✓

@jun_song · joined 04 Sep 2024

Building @0xSupergemma

43 062Followers
333Following
6 728Posts total
567.6KViews on collected posts

Latest posts

Daily reminder: Local AI has no weekly usage limits. 965 views · 31 likes · 0 reposts · 7 replies Open on X →
Never get an annual subscription. 1.1K views · 19 likes · 1 reposts · 2 replies Open on X →
I'm looking to spend about $1,200 on an annual subscription. Any recommendations? 2.8K views · 12 likes · 0 reposts · 14 replies Open on X →
It’s time for people around the world to come together and clearly define what AGI actually is. 4.9K views · 88 likes · 5 reposts · 36 replies Open on X →
GLM coding plan got much better with speed and usage right now. 14.5K views · 137 likes · 3 reposts · 8 replies Open on X →
@jun_song Really? Life has always been like this: cheap products had weak functionality, premium products were unattainable for 95% of people. You don't compare products like a Rolls Royce Phantom for 500k and a Kia Sportage. You even have a free version of QWEN 3.8. This market 1.3K views · 4 likes · 0 reposts · 0 replies Open on X →
@jun_song You didn't get the joke with "Astra". It's an allusion to asterisk, which links to disclaimers and other fine-prints. This is how it should be read: AGI* 4.3K views · 164 likes · 12 reposts · 5 replies Open on X →
@jun_song can't wait for the model named after a greek god that costs more than my rent 1.5K views · 17 likes · 0 reposts · 1 replies Open on X →
Now I get why OpenAI switched from clear names like Flash and Pro to abstract ones like Sol and Astra. (Same with Anthropic) It gives them room to stack a new name on top with every major update and justify a massive price hike. If Pro is already your flagship tier, that 51.3K views · 591 likes · 12 reposts · 34 replies Open on X →
@jun_song Almost feels like stretching the laws of physics. How can so small model outperform Mythos (on some benchmarks) which is probably 10-20x bigger. 🤯 10.8K views · 19 likes · 0 reposts · 4 replies Open on X →
@jun_song Here is the tell: it’s cheap even on public inference providers hosted in the USA. Deepseek doesn’t subsidize those afaik. Correct me if I’m wrong. 839 views · 15 likes · 0 reposts · 1 replies Open on X →
@jun_song What is funny about your take is that, for your argument to be valid, you need to have knowledge of A\ Opus’s architecture and parameter size. And it’s something I’ll bet you don’t have any knowledge of 22.9K views · 55 likes · 1 reposts · 4 replies Open on X →
People keep asking why DeepSeek’s API is so cheap. Some even make absurd claims that they’re dumping prices to corner the market. No, the answer is simple: their model size is ridiculously small compared to its performance. It's 10x smaller than Opus, and 5x smaller than 450.5K views · 3.3K likes · 197 reposts · 130 replies Open on X →

Against accounts of the same size

13 posts from the last 90 days, next to the 10K–100K follower range. shown to more people than peers of the same size.

Median views4 274this account924median for 10K–100K
Reach, %9.93%this account3.62%median for 10K–100K
Engagement, %1.21%this account1.52%median for 10K–100K
MetricThis accountMedian for 10K–100KRatio
Median views per post4 2749244.63×
Reach (views ÷ followers)9.93%3.62%2.74×
Engagement rate1.21%1.52%0.79×

Others in this range →   Compare with another account →   How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

450.5K1 Aug
22.9K
839
10.8K
51.3K3 Sep
1.5K
4.3K
1.3K
14.5K4 Sep
4.9K5 Sep
2.8K
1.1K
965

Last 13 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.81%1 Aug
0.26%
1.91%
0.21%
1.25%3 Sep
1.20%
4.23%
0.30%
1.02%4 Sep
2.66%5 Sep
0.97%
2.09%
3.94%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes81.9%4 420 in total
Reposts4.3%231 in total
Replies4.6%246 in total
Quotes0.7%40 in total
Bookmarks8.5%460 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Similar accounts