A chinese startup is giving away 100M tokens and 80 hours of A100 GPU time for free.
Intern-AI. Free sign-up, no card, no catch found yet. You get: -> ~100M tokens (10 points, 1 point ~= 20M tokens) on GLM-5.3, Qwen3.8-27B, Atria Dawn, Kimi-2.6, DeepSeek V4 (Pro/Flash) ->
A $0.07 AI model is building real tools from one prompt.
DeepSeek V4.1 Flash is showing you don't always need the biggest model for everyday work. I gave it one task: Build an SEO calculator. It returned: → One clean HTML file. → Working inputs. → Traffic projections.
Google just mogged this tame impala test.
🌀 I gave 4 fast models the same Currents cover. Same prompt. One shot via aimlapi. • Gemini 3.8 Flash 2:58 • DeepSeek V4.1 Flash 2:54 • Muse Spark 1.3 3:34 • Qwen 3.8 Omni-Flash 3:50 Four completely different interpretations.
Constantly amazed at what you can do at home with *significant* investment (but also not boat money).
My own agent/harness running entirely on my own hardware. It's the Qwen 3.8 Flash model running on one Spark, plus all the voice models running on a 3090. It's plugged into OMP,
You have reached the end of the archive
All of deepseek