Can my GPU run it?
For each GPU we have a dated price for, the two largest models (Q4_K_M) that fit in its VRAM. Other combinations are on each GPU page.
- NVIDIA GeForce RTX 5090: Qwen2.5-Coder-32B-Instruct, Qwen3-32B
- NVIDIA GeForce RTX 5080: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- NVIDIA GeForce RTX 5070 Ti: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- NVIDIA GeForce RTX 5070: phi-4, Qwen3-14B
- NVIDIA GeForce RTX 5060 Ti 16 GB: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- NVIDIA GeForce RTX 5060 Ti 8 GB: Qwen3-8B, Llama-3.1-8B-Instruct
- NVIDIA GeForce RTX 5060: Qwen3-8B, Llama-3.1-8B-Instruct
- NVIDIA GeForce RTX 4090: Qwen2.5-Coder-32B-Instruct, Qwen3-32B
- NVIDIA GeForce RTX 4070 SUPER: phi-4, Qwen3-14B
- NVIDIA GeForce RTX 3090: Qwen2.5-Coder-32B-Instruct, Qwen3-32B
- NVIDIA GeForce RTX 3060 12 GB: phi-4, Qwen3-14B
- AMD Radeon RX 9070 XT: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- AMD Radeon RX 9070: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- AMD Radeon RX 9060 XT 16 GB: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- AMD Radeon RX 7900 XTX: Qwen2.5-Coder-32B-Instruct, Qwen3-32B
- AMD Radeon RX 7800 XT: Mistral-Small-3.2-24B-Instruct-2506, gpt-oss-20b
- Intel Arc B580: phi-4, Qwen3-14B
- NVIDIA RTX A6000: Llama-3.3-70B-Instruct, Qwen2.5-Coder-32B-Instruct