Component Selection Specs Price Where Actions
Prices are dated snapshots: GPUs at list price (MSRP) unless marked used; CPUs, boards and SSDs from Newegg and PSUs, cases and RAM from Corsair on 2026-10-23 (new CPU platforms and RM650e priced 2026-11-05). We may earn a commission when you buy through these links, at no extra cost to you.
What this build runs Workload Kimi-K3.1-Pro Kimi-K3 Qwen3.8-2.4T-A95B LongCat-2.5 LongCat-2.0 DeepSeek-V4-Pro-0813 DeepSeek-V4-Pro Kimi-K2.5 Kimi-K2.6 Kimi-K2.7-Code Kimi-K2-Thinking Kimi-K2-Instruct-0905 Kimi-K2-Instruct MiMo-V2.6-Pro-RL MiMo-V2.6-Pro-MOPD MiMo-V2.5-Pro Ring-2.5-1T Ling-2.5-1T Ling-1T Ring-1T Inkling Intern-S1-Pro Muse-Spark-Open-816B-A42B Hy4-preview GLM-5 GLM-5.1 GLM-5.3 GLM-5.2 GLM-5.3-BF16 K-EXAONE-2.0-750B-A37B DeepSeek-V3.2-Exp DeepSeek-V3.2-Speciale DeepSeek-V3.2 DeepSeek-V3-0324 DeepSeek-R1-0528 DeepSeek-V3.1-Base DeepSeek-R1 DeepSeek-R1-Zero Step-5-Preview Ling-3.1-flash Ling-3.1-flash-1019 LongCat-Flash-Chat LongCat-Flash-Thinking LongCat-Flash-Thinking-2601 NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 Beam-1.0-501B-A23B Qwen3-Coder-480B-A35B-Instruct MiniMax-M1-40k MiniMax-M1-80k MiniMax-M3 ERNIE-4.5-VL-424B-A47B-PT Intern-S2-397B Qwen3.5-397B-A17B Llama-4-Maverick-17B-128E-Instruct Trinity-Large-Thinking Trinity-Large-Preview Trinity-Large-TrueBase Trinity-Large-Base Nex-N2.5-Pro Nex-N2-Pro NorthCUA-Pro GLM-4.5 GLM-4.7 GLM-4.6 GLM-5.3-Flash Motif-3 MiMo-V2.5 MiMo-V2.6-Flash-RL MiMo-V2.6-Flash-MOPD MiMo-V2-Flash DeepSeek-V4-Flash-Vision-Exp DeepSeek-V4-Flash-0731 ERNIE-4.5-300B-A47B-PT Hy3 Hy3-preview Ling-plus DeepSeek-V4-Flash MemT-1.0-Flash ERNIE-5.5-Flash-Base Inkling-Small Llama-3_1-Nemotron-Ultra-253B-v1 Solar-Open2-250B Intern-S1 K-EXAONE-236B-A23B DeepSeek-V2.5 DeepSeek-Coder-V2-Instruct DeepSeek-V2.5-1210 DeepSeek-V2-Chat-0628 DeepSeek-Coder-V2-Instruct-0724 Qwen3-VL-235B-A22B-Instruct Qwen3-235B-A22B Qwen3-235B-A22B-Instruct-2507 Qwen3-235B-A22B-Thinking-2507 MiniMax-M2 MiniMax-M2.5 MiniMax-M2.7 MiniMax-M2.1 command-a-plus-05-2026-bf16 Step-3.7-Flash Step-3.5-Flash Step-3.5-Flash-Base-Midtrain Qwen3.8-Flash-Next Mixtral-8x22B-Instruct-v0.1 Ling-3.0-flash-Sante Ling-3.0-flash Ling-3.0-flash-Fin Ornith-1.5-122B Qwen3.5-122B-A10B Devstral-2-123B-Instruct-2512 Ling-3.0-flash-VL NVIDIA-Nemotron-3-Super-120B-A12B-BF16 Mistral-Small-4-119B-2603 Youtu-LLM-A7B-Base Laguna-S-2.1 gpt-oss-120b c4ai-command-a-03-2025 GLM-4.5-Air Llama-4-Scout-17B-16E-Instruct Llama-4-Scout-17B-16E GLM-4.5V GLM-4.6V Ling-2.6-flash sarvam-105b Ling-flash-2.0 LLaDA2.1-flash Ring-flash-2.0 LLaDA2.2-flash Solar-Open-100B Qwen3-Next-80B-A3B-Instruct Qwen3-Next-80B-A3B-Thinking AliceAI-Foundation-80B-A3B-Base Hunyuan-A13B-Instruct Qwen3-Coder-Next Kolibri-1 Qwen2.5-VL-72B-Instruct UI-TARS-72B-DPO Qwen2.5-72B-Instruct Apertus-70B-Instruct-2509 Apertus-70B-2509 Llama-3.3-70B-Instruct DeepSeek-R1-Distill-Llama-70B Llama-3.1-70B Hermes-2-Theta-Llama-3-70B Meta-Llama-3.1-70B-Instruct LongCat-Flash-Lite AI21-Jamba2-Mini Llama-3_1-Nemotron-51B-Instruct Llama-3_3-Nemotron-Super-49B-v1_5 Llama-3_3-Nemotron-Super-49B-v1 Phi-3.5-MoE-instruct GRIN-MoE ALIA-40b IQuest-Coder-V1-40B-Loop-Instruct IQuest-Coder-V1-40B-Instruct Recursive Flint K2-Horizon-MoVA-36B-A4B Seed-OSS-36B-Instruct Seed-OSS-36B-Base-woSyn Intern-S2-Preview Ornith-1.5-35B-A3B Sage-A1.5 Qwen3.6-35B-A3B Qwen3.5-35B-A3B Apex-35B-V1.0 MiniCPM-V-4.7-35B-A3B MemT-1.0-Nano Agents-A1 Nex-N2.5-mini Nex-N2-mini Axion-1.0-35B NorthCUA-Mini Clover-35B-A3B-Preview Qwen-AgentWorld-35B-A3B KAT-Coder-V2.5-Dev EXAONE-4.5-33B Falcon-H1-34B-Instruct Qwen2.5-VL-32B-Instruct Laguna-XS.2 Qwen3-VL-32B-Instruct HyperCLOVAX-SEED-Think-32B FlexOlmo-7x7B-1T GLM-Z1-Rumination-32B-0414 Nanbeige4.3-33B Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 Qwen2.5-Coder-32B-Instruct DeepSeek-R1-Distill-Qwen-32B Qwen2.5-32B-Instruct QwQ-32B QwQ-32B-Preview Qwen2.5-32B Qwen2.5-Coder-32B Qwen3-32B GLM-Z1-32B-0414 GLM-4-32B-0414 OLMo-2-0325-32B-Instruct OLMo-2-0325-32B Olmo-3-32B-Think Olmo-3-1125-32B Olmo-3.1-32B-Instruct Olmo-3.1-32B-Think Olmo-3.1-32B-Instruct-DPO granite-4.0-h-small sarvam-30b EXAONE-4.0-32B EXAONE-4.0.1-32B EXAONE-3.5-32B-Instruct EXAONE-Deep-32B NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 Nemotron-Cascade-2-30B-A3B Gemma 4 31B gemma-4-31B GLM-4.7-Flash Xing4.0-29B-A4B Xing4.0-29B-Intent Qwen3-VL-30B-A3B-Instruct Qwen3-VL-30B-A3B-Thinking UI-Venus-1.5-30B-A3B kanana-2-30b-a3b-instruct-2601 kanana-2-30b-a3b-instruct kanana-2-30b-a3b-thinking-2601 Qwen3-30B-A3B Qwen3-Coder-30B-A3B-Instruct Qwen3-30B-A3B-Instruct-2507 Qwen3-30B-A3B-Thinking-2507 Tongyi-DeepResearch-30B-A3B Rubicon-Preview North-Mini-Code-1.0 Hy-Memory-30B-A3B Hy-MT2-30B-A3B granite-switch-4.1-30b Muse-Glimmer-30B ERNIE-4.5-VL-28B-A3B-Thinking ERNIE-4.5-VL-28B-A3B-PT granite-4.2-30b granite-4.1-30b Raon-1 mini preview Qwen3.8-27B Tess-4.5-27B Qwen3.6-27B Qwen3.5-27B gemma-3-27b-it Arete-27B gemma-2-27b-it gemma-2-27b Qwen3.8-27B-Longbook27-v2 medgemma-27b-text-it Hemmingway-1.5 Trinity-Mini diffusiongemma-26B-A4B-it Gemma 4 26B A4B gemma-4-26B-A4B Voxtral-Small-24B-2507 Mistral-Small-3.2-24B-Instruct-2506 Mistral-Small-3.1-24B-Instruct-2503 LFM2.5-24B-A2B LFM2-24B-A2B Mistral-Small-24B-Instruct-2501 EuroLLM-22B-Instruct-2512 EuroLLM-22B-2512 solar-pro-preview-instruct ERNIE-4.5-21B-A3B-PT ERNIE-4.5-21B-A3B-Thinking gpt-oss-20b reka-flash-3 maple-preview internlm2_5-20b-chat Groot-19B-A3B-261023 Ling-Coder-lite Ring-lite-distill-preview Ling-lite Ling-lite-1.5-2507 Ling-lite-1.5 Ring-lite Ring-mini-linear-2.0 Kimi-VL-A3B-Instruct Kimi-VL-A3B-Thinking Kimi-VL-A3B-Thinking-2506 LLaDA2.1-mini LLaDA2.0-mini Ling-mini-2.0 Ring-mini-2.0 LLaDA2.2-mini LLaDA2.0-mini-preview Moonlight-16B-A3B-Instruct Moonlight-16B-A3B Instella-MoE-16B-A3B-Think DeepSeek-Coder-V2-Lite-Instruct kanana-1.5-15.7b-a3b-instruct Phi-4-reasoning-vision-15B Apriel-Nemotron-15b-Thinker Apriel-1.5-15b-Thinker Apriel-1.6-15b-Thinker Carbon-X (DNA) Qwen2.5-Coder-14B-Instruct DeepSeek-R1-Distill-Qwen-14B Qwen2.5-14B-Instruct Qwen2.5-14B Qwen2.5-14B-Instruct-1M Qwen2.5-Coder-14B Qwen3-14B JanusCoder-14B HyperCLOVAX-SEED-Think-14B phi-4 Phi-4-reasoning Phi-4-reasoning-plus OLMo-2-1124-13B-Instruct Mistral-Nemo-Instruct-2407 gemma-3-12b-it gemma-3-12b-pt Mellum2.1-12B-A2.5B-Thinking Mellum2-12B-A2.5B-Thinking Mellum2-12B-A2.5B-Instruct Mellum2-12B-A2.5B-Thinking-SFT Mellum2.1-12B-A2.5B-Thinking-MTP Gemma 4 12B HyperCLOVAX-SEED-Omni-8B Llama-3.2-11B-Vision-Instruct Llama-3.2-11B-Vision Falcon3-10B-Instruct GLM-4.1V-9B-Thinking GLM-4.6V-Flash GLM-4.1V-9B-Base Glyph INTELLECT-1-Instruct INTELLECT-1 ZAYA1-VL-8B SenseNova-Loop-9B-Preview Ornith-1.5-9B Qwen3.5-9B MiniCPM-SALA NorthCUA-Nano GLM-4-9B-0414 GLM-Z1-9B-0414 codegeex4-all-9b glm-4-9b-chat-hf moondream3-preview gemma-2-9b-it gemma-2-9b Gemma-SEA-LION-v3-9B-IT EuroLLM-9B-Instruct-2512 NVIDIA-Nemotron-Nano-9B-v2 ZAYA1-base Yi-Coder-9B-Chat Yi-Coder-9B internlm3-8b-instruct granite-4.1-8b granite-4.2-8b granite-switch-4.1-8b Qwen3-VL-8B-Instruct Qwen3-VL-8B-Thinking UI-Venus-1.5-8B Llama-3.1-Nemotron-Nano-VL-8B-V1 MiniCPM-V-4_5 aya-vision-8b Intern-S1-mini Nemotron-Labs-Diffusion-8B-Base LFM2.5-8B-A1B Mistral-NeMo-Minitron-8B-Base LFM2-8B-A1B rnj-1-instruct MiMo-VL-7B-RL-2508 MiMo-VL-7B-RL MiMo-Embodied-7B MiMo-VL-7B-SFT Qwen2.5-VL-7B-Instruct UI-TARS-1.5-7B Fara-7B UI-Venus-Ground-7B Qwen2-VL-7B-Instruct UI-TARS-7B-DPO UI-TARS-7B-SFT Seed-Coder-8B-Instruct Seed-Coder-8B-Reasoning Stable-DiffCoder-8B-Instruct Qwen3-8B DeepSeek-R1-0528-Qwen3-8B Qwen3-8B-Base WeDLM-8B-Instruct PaCoRe-8B MiniCPM4-8B MiniCPM4.1-8B AgentCPM-Report MiniCPM4-MCP granite-3.3-8b-instruct granite-3.1-8b-instruct granite-3.2-8b-instruct granite-3.0-8b-instruct Apertus-8B-Instruct-2509 Apertus-8B-2509 Llama-3.1-Tulu-3-8B Llama-3.1-Tulu-3-8B-DPO Llama-3.1-Tulu-3.1-8B kanana-1.5-8b-instruct-2505 Llama-3.1-8B-Instruct Llama-3.1-8B DeepSeek-R1-Distill-Llama-8B Meta-Llama-3.1-8B-Instruct L3-8B-Stheno-v3.2 Llama-3.1-Nemotron-Nano-8B-v1 Meta-Llama-3.1-8B LongWriter-llama3.1-8b Hy-MT2-7B Nemotron 3.5 Bolt E4B Gemma 4 E4B Ling-3.0-tiny MiMo-7B-RL MiMo-7B-Base MiMo-7B-RL-0530 MiMo-7B-SFT EXAONE-3.5-7.8B-Instruct EXAONE-Deep-7.8B salamandra-7b-instruct salamandra-7b salamandra-7b-instruct-2606 FastVLM-7B Molmo2-O-7B internlm2_5-7b-chat internlm2_5-7b-chat-1m Phi-mini-MoE-instruct Qwen2.5-Coder-7B-Instruct DeepSeek-R1-Distill-Qwen-7B Qwen2.5-7B-Instruct Qwen2.5-7B Qwen2-7B-Instruct Qwen2.5-Coder-7B Qwen2-7B Qwen2.5-7B-Instruct-1M Qwen2.5-Math-7B Qwen2.5-Math-7B-Instruct Marco-o1 SeaLLMs-v3-7B-Chat Arcee-Spark Falcon-H1-7B-Instruct Falcon-H1R-7B starvector-8b-im2svg Falcon3-7B-Instruct Olmo-Hybrid-7B LLaDA-MoE-7B-A1B-Base Zamba2-7B-Instruct OLMo-2-1124-7B-Instruct Olmo-3-7B-Think Olmo-3-7B-Instruct Olmo-3-1025-7B Olmo-3-7B-Instruct-SFT Olmo-3-7B-Think-SFT Olmo-3-7B-Instruct-DPO Olmo-3-7B-RL-Zero-Math Olmo-3.1-7B-RL-Zero-Math Olmo-3-7B-Think-DPO Mistral-7B-Instruct-v0.3 granite-4.0-h-tiny OLMoE-1B-7B-0924-Instruct OLMoE-1B-7B-0924 OLMoE-1B-7B-0125-Instruct OLMoE-1B-7B-0125 OLMo-7B-0724-hf granite-4.0-tiny-preview YuE-s1-7B-anneal-en-cot Trinity-Nano-Preview Trinity-Nano-Base gemma-3n-E2B-it Youtu-VL-4B-Instruct Gemma 4 E2B Molmo2-ER Apriel-5B-Instruct WeVisOne-VL-4B Mage-VL Mage-VL-Cowork Qwen3.5-4B S5-nano AFM-4.5B Llama-3.1-Minitron-4B-Width-Base Qwen3-VL-4B-Instruct CapRL-Qwen3VL-4B glm-edge-4b-chat gemma-3-4b-it medgemma-4b-it medgemma-1.5-4b-it gemma-3-4b-pt Nanbeige4.2-3B Phi-3.5-vision-instruct Spark-X2.6-4B Spark-X2.5-4B POINTS-Reader MiniCPM-V-4 Qwen3-4B Qwen3-4B-Instruct-2507 Qwen3-4B-Thinking-2507 Mellum-4b-base granite-4.0-3b-vision granite-vision-4.1-4b Nemotron Labs Dory 1.0 4B NVIDIA-Nemotron-3-Nano-4B-BF16 Nanbeige4.1-3B Nanbeige4-3B-Thinking-2511 Phi-4-mini-flash-reasoning Phi-4-mini-instruct Phi-4-mini-reasoning Nemotron-Labs-Diffusion-3B Phi-3.5-mini-instruct MediPhi-Instruct MediPhi MediPhi-Clinical MediPhi-PubMed HY-Embodied-0.5 Phi-tiny-MoE-instruct CapRL-3B HyperCLOVAX-SEED-Vision-Instruct-3B kanana-1.5-v-3b-instruct granite-4.2-3b kanana-2-3b-instruct granite-4.1-3b granite-4.0-micro granite-3.1-3b-a800m-instruct Falcon3-3B-Instruct Llama-3.2-3B-Instruct Llama-3.2-3B Hermes-3-Llama-3.2-3B DeepHermes-3-Llama-3-3B-Preview AI21-Jamba-Reasoning-3B granite-4.0-h-micro Falcon-H1-3B-Instruct d1-3B LFM2.5-VL-3B Qwen2.5-3B-Instruct Qwen2.5-Coder-3B-Instruct Qwen2.5-3B Qwen2.5-Coder-3B SmolLM3-3B AI21-Jamba2-3B granite-swash-3b-a600m granite-vision-3.2-2b granite-vision-3.1-2b-preview granite-vision-3.3-2b-chart2csv-preview LFM2.5-2.6B Zamba2-2.7B Zamba2-2.7B-instruct granite-3.0-2b-instruct gemma-2-2b-it gemma-2-2b EuroMoE-2.6B-A0.6B-Instruct-Preview Motif-2.6B LFM2-2.6B LFM2-2.6B-Exp LFM2-2.6B-Transcript granite-3.3-2b-instruct granite-3.1-2b-instruct granite-3.2-2b-instruct MiniCPM5-2B MiniCPM5-2B-SFT MiniCPM5-2B-Midtrain sarvam-1-v0.5 North-Micro-Vision-Instruct UI-TARS-2B-SFT WeVisOne-VL-2B UI-Venus-1.5-2B bitnet-b1.58-2B-4T-bf16 EXAONE-3.5-2.4B-Instruct EXAONE-Deep-2.4B kanana-1.5-2.1b-instruct-2505 Qwen3.5-2B salamandra-2b salamandra-2b-instruct SmolVLM2-2.2B-Instruct SmolVLM-Base Qwen2-VL-2B-Instruct LFM3-Omni-2B granite-swash-2b Qwen3-VL-2B-Instruct kanana-nano-2.1b-instruct Hy-MT2-1.8B Qwen3-1.7B Youtu-LLM-2B LFM3-VL-2B d1-2B FastVLM-1.5B internlm2_5-1_8b-chat DeepSeek-R1-Distill-Qwen-1.5B LFM3-1.7B Qwen3-1.7B-Base SmolLM2-1.7B SmolLM-1.7B Spark-X2.6-1.7B Falcon3-1B-Instruct EuroLLM-1.7B-Instruct granite-4.0-1b LFM2.5-VL-1.6B glm-edge-1.5b-chat LFM2-VL-1.6B Falcon-H1-1.5B-Deep-Instruct Falcon-H1-1.5B-Instruct MiniCPM-V-4.7-1B Qwen2.5-1.5B-Instruct Qwen2.5-1.5B Qwen2.5-Coder-1.5B-Instruct Qwen2.5-Coder-1.5B Qwen2.5-Math-1.5B Qwen2-1.5B-Instruct Qwen2.5-Math-1.5B-Instruct ReaderLM-v2 reader-lm-1.5b Hymba-1.5B-Instruct OLMo-2-0425-1B OLMo-2-0425-1B-Instruct Yi-Coder-1.5B-Chat Bolmo-1B granite-4.0-h-1b granite-3.1-1b-a400m-instruct granite-3.0-1b-a400m-instruct MiniCPM-V-4.6 MiniCPM-V-4.6-Thinking kanana-2-1.3b-instruct OLMo-1B-0724-hf EXAONE-4.0-1.2B Llama-3.2-1B-Instruct Llama-3.2-1B Llama-3.2-1B Zamba2-1.2B-instruct HRM-Text-1B AMD-OLMo-1B LFM2.5-1.2B-Instruct LFM2-1.2B LFM2.5-1.2B-Thinking LFM2.5-1.2B-JP-202606 MiniCPM5-1B MiniCPM5-1B-SFT gemma-3-1b-it gemma-3-1b-pt Qwen3.5-0.8B Falcon-E-3B-Instruct bitnet-b1.58-2B-4T FastVLM-0.5B Qwen3-0.6B LFM2-700M Falcon-H1-Tiny-R-0.6B Qwen3-0.6B-Base d1-omni-600M HyperCLOVAX-SEED-Text-Instruct-0.5B Falcon-H1-0.5B-Instruct Qwen2.5-0.5B-Instruct Qwen2.5-0.5B Qwen2-0.5B-Instruct Qwen2.5-Coder-0.5B-Instruct LFM2-VL-450M LFM2.5-VL-450M Apertus-v1.1-0.5B MiniCPM4-0.5B circuit-sparsity SmolLM2-360M SmolLM-360M ERNIE-4.5-0.3B-PT LFM2.5-350M LFM2-350M granite-4.0-350m granite-4.0-h-350m Quant Q4_K_M Q8_0 FP16 Chat (Qwen3-14B) Coding agent (Qwen3-Coder-30B-A3B) 70B (Llama 3.3 70B)
Every model on this build (656) Model Q4_K_M Q8_0 FP16 Est. tok/s at Q4
Memory = weights + 15% + 1 GB at about 4k context; longer context adds the KV cache from each model's config (FP16). Speed = bandwidth ÷ weight size × an efficiency band calibrated on llama.cpp scoreboards; mixture-of-experts models get only an upper bound ("up to"). Multi-GPU adds memory, but speed is about one card's. How we calculate .