Quick Answer: NVIDIA H200 SXM 141GB offers 141GB VRAM and starts around current market pricing. It delivers approximately 1034 tokens/sec on Deepseek AI Deepseek Ocr 2. It typically draws 700W under load.
This GPU offers reliable throughput for local AI workloads. Pair it with the right model quantization to hit your desired tokens/sec, and monitor prices below to catch the best deal.
With 141GB VRAM, NVIDIA H200 SXM 141GB can run models up to approximately 352B parameters using 4-bit quantization. This handles most popular models including Llama 3 70B, Mistral 7B, and larger.
Consider H100 or MI300X — Maximum VRAM for enterprise workloads.
Showing 80 of 80 benchmark rows.
| Model | Size | Quantization | Tokens/sec | VRAM used |
|---|---|---|---|---|
| Deepseek AI Deepseek Ocr 2 | Unknown | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Deepseek AI Deepseek Math V2 | Unknown | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Deepseek AI Deepseek V2 5 | Unknown | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Deepseek AI Deepseek V3 | Unknown | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Deepseek AI Deepseek Coder V2 Lite Instruct | Unknown | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Deepseek AI Deepseek V3.1 | Unknown | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Deepseek AI Deepseek Coder 1.3B Instruct | 1.3B | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Deepseek AI Deepseek R1 Distill Qwen 1.5B | 1.5B | Q4 | 1033.90 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Deepseek AI Deepseek Ocr | Unknown | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 8bit | 8B | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 4bit | 8B | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Deepseek AI Deepseek R1 | Unknown | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Deepseek AI Deepseek R1 0528 | Unknown | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Deepseek AI Deepseek R1 Distill Llama 8B | 8B | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Deepseek AI Deepseek R1 Distill Qwen 7B | 7B | Q4 | 861.58 tok/sEstimated Static estimation (DB-independent) | 4GB |
| Nineninesix Kani Tts 2 En | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Nanbeige Nanbeige4 1 3B | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Minimaxai Minimax M2 5 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Minimaxai Minimax M2 1 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Stepfun AI Step 3 5 Flash | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Qwen Qwen3 Coder Next | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Moonshotai Kimi K2 5 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Xiaomimimo Mimo V2 Flash | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Nari Labs Dia2 2B | 2B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Google Embeddinggemma 300M | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Facebook Sam3 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Black Forest Labs Flux 2 Dev | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Moonshotai Kimi K2 Thinking | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Microsoft Phi 3 5 Mini Instruct | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Meta Llama Llama 3 2 3B Instruct | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Qwen Qwen3 1.7B Base | 1.7B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Dicta Il Dictalm2.0 Instruct | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2 0.5B Instruct | 0.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Apple Openelm 1 1B Instruct | 1B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2.5 3B | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Unsloth Gemma 3 1B It | 1B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Bigcode Starcoder2 3B | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Ibm Granite Granite Docling 258M | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Skt Kogpt2 Base V2 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Google Gemma 3 270M It | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Eleutherai Pythia 70M Deduped | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Microsoft Vibevoice 1.5B | 1.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Ibm Granite Granite 3.3 2B Instruct | 2B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Google Gemma 2B | 2B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Trl Internal Testing Tiny Llamaforcausallm 3.2 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Llamafactory Tiny Random Llama 3 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Unsloth Llama 3.2 1B Instruct | 1B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Numind Nuextract 1.5 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Hmellor Tiny Random Llamaforcausallm | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Sshleifer Tiny Gpt2 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Openai Community Gpt2 Xl | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Ibm Research Powermoe 3B | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Unsloth Llama 3.2 3B Instruct | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Meta Llama Llama 3.2 3B | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Eleutherai Gpt Neo 125M | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Meta Llama Llama Guard 3 1B | 1B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2 1.5B Instruct | 1.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Google Gemma 2 2B It | 2B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Microsoft Phi 3.5 Mini Instruct | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Microsoft Phi 3.5 Vision Instruct | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Rinna Japanese Gpt Neox Small | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2.5 Coder 1.5B | 1.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Microsoft Dialogpt Small | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen3 0.6B Base | 0.6B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Openai Community Gpt2 Medium | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Trl Internal Testing Tiny Random Llamaforcausallm | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2.5 Math 1.5B | 1.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Huggingfacetb Smollm 135M | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Liquidai Lfm2 1.2B | 1.2B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2 0.5B | 0.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Minimaxai Minimax M2 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Huggingfacetb Smollm2 135M | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Microsoft Phi 2 | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2.5 0.5B | 0.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen2.5 1.5B | 1.5B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Qwen Qwen3 Reranker 0.6B | 0.6B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Google T5 T5 3B | 3B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 2GB |
| Qwen Qwen3 1.7B | 1.7B | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
| Openai Community Gpt2 Large | Unknown | Q4 | 827.12 tok/sEstimated Static estimation (DB-independent) | 1GB |
Showing 240 of 240 compatibility rows.
| Model | Size | Quantization | Verdict | Estimated speed | VRAM needed |
|---|---|---|---|---|---|
| 01 AI Yi 1 5 34B Chat | 34B | Q4 | Fits comfortably | 241.24 tok/sEstimated | 17GB (have 141GB) |
| 01 AI Yi 1 5 34B Chat | 34B | Q8 | Fits comfortably | 168.87 tok/sEstimated | 34GB (have 141GB) |
| 01 AI Yi 1 5 34B Chat | 34B | FP16 | Fits comfortably | 91.67 tok/sEstimated | 68GB (have 141GB) |
| AI Forever Rugpt 3.5 13B | 13B | Q4 | Fits comfortably | 516.95 tok/sEstimated | 7GB (have 141GB) |
| AI Forever Rugpt 3.5 13B | 13B | Q8 | Fits comfortably | 361.86 tok/sEstimated | 13GB (have 141GB) |
| AI Forever Rugpt 3.5 13B | 13B | FP16 | Fits comfortably | 196.44 tok/sEstimated | 26GB (have 141GB) |
| AI Mo Kimina Prover 72B | 72B | Q4 | Fits comfortably | 137.85 tok/sEstimated | 36GB (have 141GB) |
| AI Mo Kimina Prover 72B | 72B | Q8 | Fits comfortably | 96.50 tok/sEstimated | 72GB (have 141GB) |
| AI Mo Kimina Prover 72B | 72B | FP16 | Not supported | 52.38 tok/sEstimated | 144GB (have 141GB) |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Alibaba Nlp Gte Qwen2 1.5B Instruct | 1.5B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 3GB (have 141GB) |
| Allenai Olmo 2 0425 1B | 1B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Allenai Olmo 2 0425 1B | 1B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Allenai Olmo 2 0425 1B | 1B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 2GB (have 141GB) |
| Allenai Olmo 3 7B Think | 7B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Allenai Olmo 3 7B Think | 7B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Allenai Olmo 3 7B Think | 7B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Apple Openelm 1 1B Instruct | 1B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Apple Openelm 1 1B Instruct | 1B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Apple Openelm 1 1B Instruct | 1B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 3GB (have 141GB) |
| Baichuan Inc Baichuan M2 32B | 32B | Q4 | Fits comfortably | 241.24 tok/sEstimated | 16GB (have 141GB) |
| Baichuan Inc Baichuan M2 32B | 32B | Q8 | Fits comfortably | 168.87 tok/sEstimated | 32GB (have 141GB) |
| Baichuan Inc Baichuan M2 32B | 32B | FP16 | Fits comfortably | 91.67 tok/sEstimated | 64GB (have 141GB) |
| Bigcode Starcoder2 3B | 3B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Bigcode Starcoder2 3B | 3B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Bigcode Starcoder2 3B | 3B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Bigscience Bloomz 560M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Bigscience Bloomz 560M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Bigscience Bloomz 560M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 2GB (have 141GB) |
| Black Forest Labs Flux 1 Dev | Unknown | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Black Forest Labs Flux 1 Dev | Unknown | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Black Forest Labs Flux 1 Dev | Unknown | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Black Forest Labs Flux 2 Dev | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Black Forest Labs Flux 2 Dev | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Black Forest Labs Flux 2 Dev | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 4GB (have 141GB) |
| Bsc Lt Salamandrata 7B Instruct | 7B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Bsc Lt Salamandrata 7B Instruct | 7B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Bsc Lt Salamandrata 7B Instruct | 7B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Codellama Codellama 34B HF | 34B | Q4 | Fits comfortably | 241.24 tok/sEstimated | 17GB (have 141GB) |
| Codellama Codellama 34B HF | 34B | Q8 | Fits comfortably | 168.87 tok/sEstimated | 34GB (have 141GB) |
| Codellama Codellama 34B HF | 34B | FP16 | Fits comfortably | 91.67 tok/sEstimated | 68GB (have 141GB) |
| Context Labs Meta Llama Llama 3.2 3B Instruct FP16 | 3B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Context Labs Meta Llama Llama 3.2 3B Instruct FP16 | 3B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Context Labs Meta Llama Llama 3.2 3B Instruct FP16 | 3B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Deepseek AI Deepseek Coder 1.3B Instruct | 1.3B | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 1GB (have 141GB) |
| Deepseek AI Deepseek Coder 1.3B Instruct | 1.3B | Q8 | Fits comfortably | 723.73 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek Coder 1.3B Instruct | 1.3B | FP16 | Fits comfortably | 392.88 tok/sEstimated | 3GB (have 141GB) |
| Deepseek AI Deepseek Coder 33B Instruct | 33B | Q4 | Fits comfortably | 301.55 tok/sEstimated | 17GB (have 141GB) |
| Deepseek AI Deepseek Coder 33B Instruct | 33B | Q8 | Fits comfortably | 211.09 tok/sEstimated | 33GB (have 141GB) |
| Deepseek AI Deepseek Coder 33B Instruct | 33B | FP16 | Fits comfortably | 114.59 tok/sEstimated | 66GB (have 141GB) |
| Deepseek AI Deepseek Coder V2 Instruct 0724 | Unknown | Q4 | Fits comfortably | 172.32 tok/sEstimated | 36GB (have 141GB) |
| Deepseek AI Deepseek Coder V2 Instruct 0724 | Unknown | Q8 | Fits comfortably | 120.62 tok/sEstimated | 72GB (have 141GB) |
| Deepseek AI Deepseek Coder V2 Instruct 0724 | Unknown | FP16 | Not supported | 65.48 tok/sEstimated | 144GB (have 141GB) |
| Deepseek AI Deepseek Coder V2 Lite Instruct | Unknown | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 1GB (have 141GB) |
| Deepseek AI Deepseek Coder V2 Lite Instruct | Unknown | Q8 | Fits comfortably | 723.73 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek Coder V2 Lite Instruct | Unknown | FP16 | Fits comfortably | 392.88 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek Math V2 | Unknown | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 1GB (have 141GB) |
| Deepseek AI Deepseek Math V2 | Unknown | Q8 | Fits comfortably | 723.73 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek Math V2 | Unknown | FP16 | Fits comfortably | 392.88 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek Ocr | Unknown | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek Ocr | Unknown | Q8 | Fits comfortably | 603.11 tok/sEstimated | 7GB (have 141GB) |
| Deepseek AI Deepseek Ocr | Unknown | FP16 | Fits comfortably | 327.40 tok/sEstimated | 14GB (have 141GB) |
| Deepseek AI Deepseek Ocr 2 | Unknown | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 1GB (have 141GB) |
| Deepseek AI Deepseek Ocr 2 | Unknown | Q8 | Fits comfortably | 723.73 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek Ocr 2 | Unknown | FP16 | Fits comfortably | 392.88 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek R1 | Unknown | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek R1 | Unknown | Q8 | Fits comfortably | 603.11 tok/sEstimated | 7GB (have 141GB) |
| Deepseek AI Deepseek R1 | Unknown | FP16 | Fits comfortably | 327.40 tok/sEstimated | 14GB (have 141GB) |
| Deepseek AI Deepseek R1 0528 | Unknown | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek R1 0528 | Unknown | Q8 | Fits comfortably | 603.11 tok/sEstimated | 8GB (have 141GB) |
| Deepseek AI Deepseek R1 0528 | Unknown | FP16 | Fits comfortably | 327.40 tok/sEstimated | 16GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Llama 8B | 8B | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Llama 8B | 8B | Q8 | Fits comfortably | 603.11 tok/sEstimated | 8GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Llama 8B | 8B | FP16 | Fits comfortably | 327.40 tok/sEstimated | 16GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 1.5B | 1.5B | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 1GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 1.5B | 1.5B | Q8 | Fits comfortably | 723.73 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 1.5B | 1.5B | FP16 | Fits comfortably | 392.88 tok/sEstimated | 3GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 32B | 32B | Q4 | Fits comfortably | 301.55 tok/sEstimated | 16GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 32B | 32B | Q8 | Fits comfortably | 211.09 tok/sEstimated | 32GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 32B | 32B | FP16 | Fits comfortably | 114.59 tok/sEstimated | 64GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 7B | 7B | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 7B | 7B | Q8 | Fits comfortably | 603.11 tok/sEstimated | 7GB (have 141GB) |
| Deepseek AI Deepseek R1 Distill Qwen 7B | 7B | FP16 | Fits comfortably | 327.40 tok/sEstimated | 14GB (have 141GB) |
| Deepseek AI Deepseek V2 5 | Unknown | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 1GB (have 141GB) |
| Deepseek AI Deepseek V2 5 | Unknown | Q8 | Fits comfortably | 723.73 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek V2 5 | Unknown | FP16 | Fits comfortably | 392.88 tok/sEstimated | 4GB (have 141GB) |
| Deepseek AI Deepseek V3 | Unknown | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek V3 | Unknown | Q8 | Fits comfortably | 723.73 tok/sEstimated | 3GB (have 141GB) |
| Deepseek AI Deepseek V3 | Unknown | FP16 | Fits comfortably | 392.88 tok/sEstimated | 6GB (have 141GB) |
| Deepseek AI Deepseek V3 0324 | Unknown | Q4 | Fits comfortably | 301.55 tok/sEstimated | 16GB (have 141GB) |
| Deepseek AI Deepseek V3 0324 | Unknown | Q8 | Fits comfortably | 211.09 tok/sEstimated | 32GB (have 141GB) |
| Deepseek AI Deepseek V3 0324 | Unknown | FP16 | Fits comfortably | 114.59 tok/sEstimated | 64GB (have 141GB) |
| Deepseek AI Deepseek V3.1 | Unknown | Q4 | Fits comfortably | 1033.90 tok/sEstimated | 2GB (have 141GB) |
| Deepseek AI Deepseek V3.1 | Unknown | Q8 | Fits comfortably | 723.73 tok/sEstimated | 3GB (have 141GB) |
| Deepseek AI Deepseek V3.1 | Unknown | FP16 | Fits comfortably | 392.88 tok/sEstimated | 6GB (have 141GB) |
| Dicta Il Dictalm2.0 Instruct | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Dicta Il Dictalm2.0 Instruct | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Dicta Il Dictalm2.0 Instruct | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 4GB (have 141GB) |
| Distilbert Distilgpt2 | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Distilbert Distilgpt2 | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Distilbert Distilgpt2 | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 4GB (have 141GB) |
| Dphn Dolphin 2.9.1 Yi 1.5 34B | 34B | Q4 | Fits comfortably | 241.24 tok/sEstimated | 17GB (have 141GB) |
| Dphn Dolphin 2.9.1 Yi 1.5 34B | 34B | Q8 | Fits comfortably | 168.87 tok/sEstimated | 34GB (have 141GB) |
| Dphn Dolphin 2.9.1 Yi 1.5 34B | 34B | FP16 | Fits comfortably | 91.67 tok/sEstimated | 68GB (have 141GB) |
| Eleutherai Gpt Neo 125M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Eleutherai Gpt Neo 125M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Eleutherai Gpt Neo 125M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Eleutherai Pythia 70M Deduped | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Eleutherai Pythia 70M Deduped | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Eleutherai Pythia 70M Deduped | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Essentialai Rnj 1 | Unknown | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Essentialai Rnj 1 | Unknown | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Essentialai Rnj 1 | Unknown | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Facebook Opt 125M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Facebook Opt 125M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Facebook Opt 125M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Facebook Sam3 | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Facebook Sam3 | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Facebook Sam3 | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Fireredteam Firered Image Edit 1 0 | Unknown | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Fireredteam Firered Image Edit 1 0 | Unknown | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Fireredteam Firered Image Edit 1 0 | Unknown | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Gensyn Qwen2.5 0.5B Instruct | 0.5B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Gensyn Qwen2.5 0.5B Instruct | 0.5B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Gensyn Qwen2.5 0.5B Instruct | 0.5B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Google Bert Bert Base Uncased | Unknown | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Google Bert Bert Base Uncased | Unknown | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Google Bert Bert Base Uncased | Unknown | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Google Embeddinggemma 300M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Google Embeddinggemma 300M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Google Embeddinggemma 300M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 2 27B It | 27B | Q4 | Fits comfortably | 379.09 tok/sEstimated | 14GB (have 141GB) |
| Google Gemma 2 27B It | 27B | Q8 | Fits comfortably | 265.37 tok/sEstimated | 27GB (have 141GB) |
| Google Gemma 2 27B It | 27B | FP16 | Fits comfortably | 144.06 tok/sEstimated | 54GB (have 141GB) |
| Google Gemma 2 2B It | 2B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 2 2B It | 2B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Google Gemma 2 2B It | 2B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 4GB (have 141GB) |
| Google Gemma 2 9B It | 9B | Q4 | Fits comfortably | 516.95 tok/sEstimated | 5GB (have 141GB) |
| Google Gemma 2 9B It | 9B | Q8 | Fits comfortably | 361.86 tok/sEstimated | 9GB (have 141GB) |
| Google Gemma 2 9B It | 9B | FP16 | Fits comfortably | 196.44 tok/sEstimated | 18GB (have 141GB) |
| Google Gemma 2B | 2B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 2B | 2B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Google Gemma 2B | 2B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 4GB (have 141GB) |
| Google Gemma 3 1B It | 1B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 3 1B It | 1B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 3 1B It | 1B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 2GB (have 141GB) |
| Google Gemma 3 270M It | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 3 270M It | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Google Gemma 3 270M It | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Google T5 T5 3B | 3B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Google T5 T5 3B | 3B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Google T5 T5 3B | 3B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Gsai Ml Llada 8B Base | 8B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Gsai Ml Llada 8B Base | 8B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 8GB (have 141GB) |
| Gsai Ml Llada 8B Base | 8B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 16GB (have 141GB) |
| Gsai Ml Llada 8B Instruct | 8B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Gsai Ml Llada 8B Instruct | 8B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 8GB (have 141GB) |
| Gsai Ml Llada 8B Instruct | 8B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 16GB (have 141GB) |
| Hmellor Tiny Random Llamaforcausallm | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Hmellor Tiny Random Llamaforcausallm | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Hmellor Tiny Random Llamaforcausallm | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Huggingfaceh4 Zephyr 7B Beta | 7B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Huggingfaceh4 Zephyr 7B Beta | 7B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Huggingfaceh4 Zephyr 7B Beta | 7B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Huggingfacem4 Tiny Random Llamaforcausallm | Unknown | Q4 | Fits comfortably | 689.26 tok/sEstimated | 2GB (have 141GB) |
| Huggingfacem4 Tiny Random Llamaforcausallm | Unknown | Q8 | Fits comfortably | 482.48 tok/sEstimated | 4GB (have 141GB) |
| Huggingfacem4 Tiny Random Llamaforcausallm | Unknown | FP16 | Fits comfortably | 261.92 tok/sEstimated | 8GB (have 141GB) |
| Huggingfacetb Smollm 135M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Huggingfacetb Smollm 135M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Huggingfacetb Smollm 135M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Huggingfacetb Smollm2 135M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Huggingfacetb Smollm2 135M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Huggingfacetb Smollm2 135M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Huggyllama Llama 7B | 7B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Huggyllama Llama 7B | 7B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Huggyllama Llama 7B | 7B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Ibm Granite Granite 3.3 2B Instruct | 2B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Ibm Granite Granite 3.3 2B Instruct | 2B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Ibm Granite Granite 3.3 2B Instruct | 2B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 4GB (have 141GB) |
| Ibm Granite Granite 3.3 8B Instruct | 8B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Ibm Granite Granite 3.3 8B Instruct | 8B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 8GB (have 141GB) |
| Ibm Granite Granite 3.3 8B Instruct | 8B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 16GB (have 141GB) |
| Ibm Granite Granite Docling 258M | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Ibm Granite Granite Docling 258M | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 1GB (have 141GB) |
| Ibm Granite Granite Docling 258M | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 1GB (have 141GB) |
| Ibm Research Powermoe 3B | 3B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Ibm Research Powermoe 3B | 3B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Ibm Research Powermoe 3B | 3B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Ilyagusev Saiga Llama3 8B | 8B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 2GB (have 141GB) |
| Ilyagusev Saiga Llama3 8B | 8B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 4GB (have 141GB) |
| Ilyagusev Saiga Llama3 8B | 8B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 8GB (have 141GB) |
| Inference Net Schematron 3B | 3B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Inference Net Schematron 3B | 3B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Inference Net Schematron 3B | 3B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Kaitchup Phi 3 Mini 4K Instruct Gptq 4bit | Unknown | Q4 | Fits comfortably | 689.26 tok/sEstimated | 2GB (have 141GB) |
| Kaitchup Phi 3 Mini 4K Instruct Gptq 4bit | Unknown | Q8 | Fits comfortably | 482.48 tok/sEstimated | 4GB (have 141GB) |
| Kaitchup Phi 3 Mini 4K Instruct Gptq 4bit | Unknown | FP16 | Fits comfortably | 261.92 tok/sEstimated | 8GB (have 141GB) |
| Liquidai Lfm2 1.2B | 1.2B | Q4 | Fits comfortably | 827.12 tok/sEstimated | 1GB (have 141GB) |
| Liquidai Lfm2 1.2B | 1.2B | Q8 | Fits comfortably | 578.98 tok/sEstimated | 2GB (have 141GB) |
| Liquidai Lfm2 1.2B | 1.2B | FP16 | Fits comfortably | 314.30 tok/sEstimated | 3GB (have 141GB) |
| Liuhaotian Llava V1.5 7B | 7B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Liuhaotian Llava V1.5 7B | 7B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Liuhaotian Llava V1.5 7B | 7B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Llamafactory Tiny Random Llama 3 | Unknown | Q4 | Fits comfortably | 827.12 tok/sEstimated | 2GB (have 141GB) |
| Llamafactory Tiny Random Llama 3 | Unknown | Q8 | Fits comfortably | 578.98 tok/sEstimated | 3GB (have 141GB) |
| Llamafactory Tiny Random Llama 3 | Unknown | FP16 | Fits comfortably | 314.30 tok/sEstimated | 6GB (have 141GB) |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 4bit | 8B | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 4bit | 8B | Q8 | Fits comfortably | 603.11 tok/sEstimated | 8GB (have 141GB) |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 4bit | 8B | FP16 | Fits comfortably | 327.40 tok/sEstimated | 16GB (have 141GB) |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 8bit | 8B | Q4 | Fits comfortably | 861.58 tok/sEstimated | 4GB (have 141GB) |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 8bit | 8B | Q8 | Fits comfortably | 603.11 tok/sEstimated | 8GB (have 141GB) |
| Lmstudio Community Deepseek R1 0528 Qwen3 8B Mlx 8bit | 8B | FP16 | Fits comfortably | 327.40 tok/sEstimated | 16GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 4bit | 4B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 2GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 4bit | 4B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 4GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 4bit | 4B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 8GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 6bit | 4B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 2GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 6bit | 4B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 4GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 6bit | 4B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 8GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 8bit | 4B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 2GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 8bit | 4B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 4GB (have 141GB) |
| Lmstudio Community Qwen3 4B Thinking 2507 Mlx 8bit | 4B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 8GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 4bit | 30B | Q4 | Fits comfortably | 379.09 tok/sEstimated | 15GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 4bit | 30B | Q8 | Fits comfortably | 265.37 tok/sEstimated | 30GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 4bit | 30B | FP16 | Fits comfortably | 144.06 tok/sEstimated | 60GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 5bit | 30B | Q4 | Fits comfortably | 379.09 tok/sEstimated | 15GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 5bit | 30B | Q8 | Fits comfortably | 265.37 tok/sEstimated | 30GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 5bit | 30B | FP16 | Fits comfortably | 144.06 tok/sEstimated | 60GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 6bit | 30B | Q4 | Fits comfortably | 379.09 tok/sEstimated | 15GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 6bit | 30B | Q8 | Fits comfortably | 265.37 tok/sEstimated | 30GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 6bit | 30B | FP16 | Fits comfortably | 144.06 tok/sEstimated | 60GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 8bit | 30B | Q4 | Fits comfortably | 379.09 tok/sEstimated | 15GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 8bit | 30B | Q8 | Fits comfortably | 265.37 tok/sEstimated | 30GB (have 141GB) |
| Lmstudio Community Qwen3 Coder 30B A3b Instruct Mlx 8bit | 30B | FP16 | Fits comfortably | 144.06 tok/sEstimated | 60GB (have 141GB) |
| Lmsys Vicuna 7B V1.5 | 7B | Q4 | Fits comfortably | 689.26 tok/sEstimated | 4GB (have 141GB) |
| Lmsys Vicuna 7B V1.5 | 7B | Q8 | Fits comfortably | 482.48 tok/sEstimated | 7GB (have 141GB) |
| Lmsys Vicuna 7B V1.5 | 7B | FP16 | Fits comfortably | 261.92 tok/sEstimated | 14GB (have 141GB) |
| Meta Llama Llama 2 13B Chat HF | 13B | Q4 | Fits comfortably | 516.95 tok/sEstimated | 7GB (have 141GB) |
| Meta Llama Llama 2 13B Chat HF | 13B | Q8 | Fits comfortably | 361.86 tok/sEstimated | 13GB (have 141GB) |
| Meta Llama Llama 2 13B Chat HF | 13B | FP16 | Fits comfortably | 196.44 tok/sEstimated | 26GB (have 141GB) |
Note: Performance estimates are calculated. Real results may vary. Methodology · Submit real data
Buy directly on Amazon with fast shipping and reliable customer service.
Essential accessories to pair with NVIDIA H200 SXM 141GB
Total Bundle Price
All items from Amazon
💡 Not ready to buy? Try cloud GPUs first
Test NVIDIA H200 SXM 141GB performance in the cloud before investing in hardware. Pay by the hour with no commitment.
Explore how RTX 4090 stacks up for local inference workloads.
Explore how RTX 4080 stacks up for local inference workloads.
Explore how RTX 4070 Ti stacks up for local inference workloads.
Explore how RTX 3090 stacks up for local inference workloads.
Explore how RX 7900 XTX stacks up for local inference workloads.
RPG • 2020
RPG • 2023
Action RPG • 2023
RPG • 2023
Survival Horror • 2023
Action RPG • 2022
Action RPG • 2024
Action Adventure • 2025
Survival Horror • 2023
Action • 2022
Action Adventure • 2023
Action Adventure • 2019