Running 5 TurboQuant on Consumer GPUs β 100K Context on RTX 3090, 64K on RTX 4070 π 5 Extend LLM context to 100K tokens on consumer GPUs
Qwen/Qwen3-Coder-30B-A3B-Instruct Text Generation β’ 31B β’ Updated Dec 3, 2025 β’ 1.67M β’ β’ 1.01k
Running on Zero Featured 260 SmolDocling π¦ 260 Convert images and queries into structured document text