Four models you would usually pay for are sitting on free endpoints right now, no card.
Sensetime token plan public beta: deepseek v4 flash and glm-5.2: 500 calls / 5 hours 6.8 flash lite and u1 fast: 1,500 calls / 5 hours talks openai and
Snowboarding game created in local environment qwen3.8-27b-mtplx-optimized-speed-fp16
I was able to create it, but it's better to start with deepseek v4 flash or GLM-5.3 flash than to start from scratch now, so I'm closing it. It's not something I would go out of my way to make locally...
Testing out DeepSeek V4 on a Mac Studio M3 Ultra for app development.
Can't wait to test out bigger models on this 256GB Machine! Stay tuned.
In China, that's exactly what happens.
Curry's brain was communist brainwashed. In China there is already an App and a Drone that women call... The problem is that if you add the cost of the AI Drone + deepseek + tencent/alibaba = +1 trillion USD Brazil is 60 years behind.
You have reached the end of the archive
All of deepseek