will it local
32GB Battlemage Pro — the big-VRAM Intel play for 32-33B-class fits at ~$1.2-1.5k street. Gaming performance is mediocre for the price (Tom's Hardware: ~2x B580); this is a VRAM-density buy, not a speed buy.
computed band 71.0-92.0 tok/s for 8B Q4_K_M (roofline, 608.0 GB/s VRAM); community: ~22 tok/s 27B dense Q4_K_M (lmx-harvest 2026-09-06); community: ~25 tok/s 33B dense Q4_K_M (lmx-harvest 2026-09-06)
Effective decode window: 0.60–0.78 of 608 GB/s nominal → ~365–474 GB/s effective (llama.cpp decode, Q4_K_M basis; arch window, community-calibrated)
Community check — localmaxxing.com medians (3 buckets): localmaxxing.com
→ ~97 tok/s · 3-4B Q4 · 3 runs · below our band — expected for small models (kernel-launch overhead)
→ ~26 tok/s · 30-33B Q4 · 16 runs · just above our band (12% over the high edge)
→ ~4.9 tok/s · 30-33B Q5 · 4 runs · below our band — expected for small models (kernel-launch overhead)
→ 13b-q4 full 14.6 GB needed of 32 GB usable — headroom for context
→ 3b-q4 full 2.8 GB needed of 32 GB usable — headroom for context
→ 4b-q4 full 3.7 GB needed of 32 GB usable — headroom for context
→ 8b-q4 full 6.1 GB needed of 32 GB usable — headroom for context
→ ~22 tok/s · 27B dense Q4_K_M localmaxxing community, n=5, GGUF llama.cpp (gemma-3-27b-class) · lmx-harvest 2026-09-06
→ ~25 tok/s · 33B dense Q4_K_M localmaxxing community, n=5, GGUF llama.cpp (eff ~0.87 — above battlemage window top, kept verbatim; single-family small-n) · lmx-harvest 2026-09-06
Nominal bandwidth: 608.0 GB/s — see the effective decode window above
→ qwen3:0.6b on Tesla M60 (16GB, used) ($35) 0.6B · 0.6GB · needs ~1GB (4k ctx)
→ qwen3:1.7b on Tesla M60 (16GB, used) ($35) 1.7B · 1.4GB · needs ~1.8GB (4k ctx)
→ qwen3:4b on Tesla M60 (16GB, used) ($35) 4B · 2.6GB · needs ~3.2GB (4k ctx)
→ llama3.2:1b on Tesla M60 (16GB, used) ($35) 1B · 1.3GB · needs ~1.4GB (4k ctx)
→ llama3.2:3b on Tesla M60 (16GB, used) ($35) 3B · 2.0GB · needs ~2.4GB (4k ctx)
→ gemma3:1b on Tesla M60 (16GB, used) ($35) 1B · 1.0GB · needs ~1.1GB (4k ctx)
→ phi4:mini on Tesla M60 (16GB, used) ($35) mini 3.8B · 2.5GB · needs ~3GB (4k ctx)
→ moondream:2b on Tesla M60 (16GB, used) ($35) 2B · 1.7GB · needs ~3.7GB (4k ctx)
→ whisper:tiny on Tesla M60 (16GB, used) ($35) tiny · 0.1GB · needs ~2.1GB (4k ctx)
→ whisper on Tesla M60 (16GB, used) ($35) base · 0.3GB · needs ~2.3GB (4k ctx)
→ whisper:small on Tesla M60 (16GB, used) ($35) small · 0.9GB · needs ~2.9GB (4k ctx)
→ whisper:medium on Tesla M60 (16GB, used) ($35) medium · 3.1GB · needs ~5.1GB (4k ctx)
→ kokoro:82m on Tesla M60 (16GB, used) ($35) 82M · 0.3GB · needs ~2.3GB (4k ctx)
→ gemma3:4b on Tesla M60 (16GB, used) ($35) 4B · 3.3GB · needs ~3.8GB (4k ctx)
→ deepseek-r1:7b on Tesla M60 (16GB, used) ($35) distill 7B · 4.7GB · needs ~4.9GB (4k ctx)
→ mistral:7b on Tesla M60 (16GB, used) ($35) 7B · 4.1GB · needs ~4.6GB (4k ctx)
→ llava:7b on Tesla M60 (16GB, used) ($35) 7B · 4.7GB · needs ~6.7GB (4k ctx)
→ qwen3:8b on Tesla M60 (16GB, used) ($35) 8B · 5.2GB · needs ~5.8GB (4k ctx)
→ qwen2.5vl:7b on Tesla M60 (16GB, used) ($35) 7B · 5.6GB · needs ~5.8GB (4k ctx)
→ llama3.1:8b on Tesla M60 (16GB, used) ($35) 8B · 4.9GB · needs ~5.4GB (4k ctx)
→ deepseek-r1:8b on Tesla M60 (16GB, used) ($35) distill 8B · 4.9GB · needs ~5.4GB (4k ctx)
→ whisper:large-v3 on Tesla M60 (16GB, used) ($35) large-v3 · 6.2GB · needs ~8.2GB (4k ctx)
→ mistral-nemo:12b on Tesla M60 (16GB, used) ($35) 12B · 7.1GB · needs ~7.7GB (4k ctx)
→ sdxl on Tesla M60 (16GB, used) ($35) SDXL base · 6.9GB · needs ~8.9GB (4k ctx)
→ llava:13b on Tesla M60 (16GB, used) ($35) 13B · 8.0GB · needs ~11.1GB (4k ctx)
→ gemma3:12b on Tesla M60 (16GB, used) ($35) 12B · 8.1GB · needs ~9.6GB (4k ctx)
→ qwen3:14b on Tesla M60 (16GB, used) ($35) 14B · 9.3GB · needs ~9.9GB (4k ctx)
→ deepseek-r1:14b on Tesla M60 (16GB, used) ($35) distill 14B · 9.0GB · needs ~9.8GB (4k ctx)
→ phi4:14b on Tesla M60 (16GB, used) ($35) 14B · 9.1GB · needs ~9.9GB (4k ctx)
→ gpt-oss:20b on Tesla M40 (24GB, used) ($50) 20B MoE · 14GB · needs ~14.2GB (4k ctx)
→ devstral:24b on Tesla M40 (24GB, used) ($50) 24B · 14.6GB · needs ~15.2GB (4k ctx)
→ qwen3.8:27b on Tesla V100 32GB (used) ($700) 27B · 16.5GB · needs ~18.5GB (4k ctx)
→ gemma3:27b on Tesla V100 32GB (used) ($700) 27B · 17GB · needs ~18.9GB (4k ctx)
→ flux:schnell on Tesla V100 32GB (used) ($700) schnell fp8 · 17GB · needs ~19GB (4k ctx)
→ qwen3:30b-a3b on Tesla V100 32GB (used) ($700) 30B-A3B MoE · 18.6GB · needs ~19GB (4k ctx)
→ qwen3-coder:30b-a3b on Tesla V100 32GB (used) ($700) 30B-A3B MoE · 18.6GB · needs ~19GB (4k ctx)
→ qwen3:32b on Tesla V100 32GB (used) ($700) 32B · 20GB · needs ~21GB (4k ctx)
→ deepseek-r1:32b on Tesla V100 32GB (used) ($700) distill 32B · 20GB · needs ~21GB (4k ctx)
→ qwen2.5vl:32b on Mac mini M5 Pro (64GB) ($1669) 32B · 22GB · needs ~23GB (4k ctx)
→ flux:dev on Mac mini M5 Pro (64GB) ($1669) dev fp8 · 23GB · needs ~25GB (4k ctx)
→ mixtral:8x7b on Mac mini M5 Pro (64GB) ($1669) 8x7B MoE · 26GB · needs ~26.5GB (4k ctx)
—
⚠ NOT 48GB — the 09-04 coverage-list row guessed 48GB '(t)'; TechPowerUp pins 32GB GDDR6 / 256-bit / 608 GB/s (list rule: trust TPU over the list)
⚠ New card (2026): community data thin and mixed — MoE/hybrid models (gemma-E4B class, Qwen3.6-35B-A3B class) post inflated rows that are NOT dense-model speeds; only GGUF dense rows are comparable to our bands
⚠ GPTQ/vLLM rows (GPTQ-Int4, sym_int4) are a different kernel family than llama.cpp GGUF — not comparable, excluded from band reasoning
⚠ No 8B row found in localmaxxing harvest — band computed from roofline; dense 28-33B GGUF rows imply eff 0.58-0.87 (wide, small n)