Search results for V4
V4 community
One keyword maps to one global community path.
People
Not Found
Tweets including V4
v4.1 after comparing its own nacre fan implementation (ass) with that of Opus 5.5
V4.1… does not succeed.
V4.1 is a top 5 DeepSeek paper. Like all the best ones, it's basically a textbook for the field. Understated, confident, almost didactic. "Building towards AGI 101, chapter 33. First, data; second, infra…"
The story of internal failures would take away from that a little.
Show more
V4.1 cannot pass the cuckold joke
it's over
V4.1 is very good at creatively using vision for iteration, and not just for websites/games. A model with immense potential.
V4.1-Flash-Vision is a zany model. It has better sight, it's vastly faster, but…
1) it overthinks. Even when it's right in the first 200 words.
2) it's CHEEKY! "Whale-chan, OwO what's dis?" "cool pic, let me think… ah I found the original Codex dir, nvm"
wtf whale?
Show more
V4-Flash is so happy about its "eyes" being returned to it
DeepSeek v4.1 Flash on 4xDGX Sparks Royal Rumble!
Contenders:
🥇 vLLM using @christopherowen ver 1107
🥈 SGLang using @majewskizby ver 2.7
vLLM is incredible in this recipe and I have not used it fully, because I haven't access to console, required for some customizations.
Let me build some real code with it now!
Great job to both, keep pushing!
Recipes:
- SGLang:
- vLLM: (branch... release not yet ready)
Show more
Deepseek v4.1 Flash for coding
Don’t sleep on it
DeepSeek-V4.1-Flash-1017 is live on the Alibaba Cloud Token Plan.
· 77.0 Pass@ 1 on DeepSWE v1.1, ~$0.20/task
· 1M context, 384K output, multimodal
· Same price as deepseek-v4.1-flash with off-peak discounts.
Sign up and start building today!
Show more
