tweetindex

antirez @antirez · 02 Sep 2026

P.S. for me DeepSeek v4 Flash is still the king of local inference, regardless of other great models released that could beat it in specific benchmarks and when many (many!) tokens are used. Especially now that DS4F has vision. What a model...
58 775Views
787Likes
31Reposts
82Replies
9Quotes
91Bookmarks

Is that a lot?

3.40×vs this author's median17 262 views is typical
75Percentile for this authorof 4 recent posts
21.4×vs 10K–100K median2 751 views is typical
73.59%Reachviews ÷ followers
1.55%Engagement rateof viewers reacted

Compare with the benchmark table →

Open on X →