llamaperf

Local model performance on your hardware

Find which open-weight LLMs fit in your GPU or Mac and compare the speeds people report on setups like yours.

What runs on your hardware?

Pick your GPU or Mac, then read the speeds people reported on it, or estimate which models fit and how fast they run.

Free to use. No account needed. Memory estimates and community measurements are labelled separately.

Looking for a particular model?

Search for a model to see its reported speeds across GPUs and Macs.

Local model performance reports from the community

These are individual setups, not a controlled benchmark. Compare GPU count, quantization, context and offloading before comparing speeds. How to read a report →

Model: Minimax M2.1
Compare setup details

Exact recorded values. Context may be a configured limit; matching filters does not establish identical prompts, offloading or concurrency.

Minimax M2.1 IQ4_XS / ROCm b7822 / native PP256 and separate TG256

AMD Radeon Graphics gfx1151 (marketing SKU and installed RAM unreported) · llama.cpp b7822 / 8f91ca54e / ROCm 7.1.1

generation:
28.9 tokens/s
prompt processing (prefill):
147.6 tokens/s
quant:
IQ4_XS (GGUF)
flash attention:
on

User reports Minimax M2.1 at 28.9 t/s generation and 147.6 t/s prompt processing on an AMD Radeon Graphics gfx1151 device. Setup is llama-bench on ROCm 7.1.1 with IQ4_XS, ngl=999, FA=1, and dev=ROCm0, selecting one GPU. The native figures are PP256 147.63 +/- 2.44 and TG256 28.92 +/- 0.21; exact context depth, CPU threads, and repetitions are unknown.

Oct 14, 2026

Minimax M2.1 IQ4_XS / ROCm b7819 / native PP4096 and separate TG512

AMD Radeon 8060S integrated GPU (Strix Halo; installed RAM unreported) · llama.cpp b7819 / 557515be1 / ROCm 6.4.4

generation:
28.3 tokens/s
prompt processing (prefill):
248.5 tokens/s
quant:
IQ4_XS (GGUF)
flash attention:
on

User reports Minimax M2.1 at 28.3 t/s generation and 248.5 t/s prompt processing on a single Radeon 8060S Graphics (gfx1151) via ROCm. Setup is llama-bench with IQ4_XS at 4.25 bpw, -ngl 999, -fa 1, -p 4096 and -n 512, one repetition. Native figures were PP4096 248.52 and TG512 28.28 t/s, both with zero deviation; the model is 113.52 GiB and 228.69 B parameters.

Oct 14, 2026
AMD hardware
generation:
31.1 tokens/s
prompt processing (prefill):
104.8 tokens/s
quant:
UD-Q3_K_XL (GGUF)
flash attention:
on

User reports Minimax M2.1 at 31.1 t/s generation and 104.8 t/s prompt processing on an AMD Strix Halo 128GB machine. Setup is llama-bench on Vulkan0 with the UD-Q3_K_XL quant, -ngl 999, -p 256 -n 256, -t 16, -r 3 and -fa 1. The native figures are PP256 104.80 +/- 7.95 and TG256 31.13 +/- 0.02; a separate toolbox run gave PP256 104.50 +/- 8.57 and TG256 33.07 +/- 0.05.

Oct 14, 2026

Minimax M2.1 230B (10B active)

2× RTX 4090D 48GB · llama.cpp · 65,536 ctx

Tone: positive
generation:
20.1 tokens/s
prompt processing (prefill):
67.1 tokens/s
quant:
UD-Q4_K_XL (GGUF)

User reports Minimax M2.1 at 20.05 t/s generation and 67.11 t/s prompt eval on an RTX 4090D 48GB plus an RTX 3090. Setup is llama.cpp build 7542 with UD-Q4_K_XL GGUF at 65536 context, running in router mode on a Xeon W5-3425 with 256GB DDR5. User notes speed is lower at larger context and says it works great.

Oct 9, 2026
Tone: negativeApple hardware
generation:
25.0 tokens/s
quant:
4bit
agenticcodinglong-context

User reports Minimax-m2.1-4bit at 25 t/s generation on an M3 Ultra 256GB with 80-core GPU, compared against llama.cpp with flash attention at 32 t/s. Setup is MLX with 4bit weights at 30K context; the same model under llama.cpp with flash attention reached 32 t/s generation and 78s prompt processing. At 146K context MLX fell to 5.95 t/s generation versus 12.12 t/s for llama.cpp, and the user says MLX runs about 50% slower on long-context generation, impacting agentic coding.

Oct 4, 2026
Showing 1–5 of 5
Page 1 of 1

Community benchmarks snapshot

Records by GPU

AMD Strix Halo 128GB348NVIDIA RTX 3090271NVIDIA RTX 5090197NVIDIA DGX Spark99NVIDIA RTX 3060 12GB96NVIDIA RTX 5060 Ti 16GB88NVIDIA RTX Pro 6000 Blackwell81AMD Radeon AI PRO R9700 32GB81NVIDIA RTX 5070 Ti77NVIDIA RTX 409068

Records by model

3266 total
Qwen3.81567
Qwen3.6334
Gemma 4185
DeepSeek V4 Flash150
Qwen3.5145
DeepSeek V4.1 Flash103
other782

Records by engine

2308 total
llama.cpp1134
Strata234
vLLM227
MLX107
Ollama99
other507

Use cases

coding 937agentic 640long-context 442tool-use 281vision 174summarization 81math 77creative-writing 58multilingual 35rp 22text-generation 9reasoning 3
coding937agentic640long-context442tool-use281vision174summarization81math77creative-writing58

Median generation t/s by GPU

On Qwen3.8 27B at 4-bit, plain single-GPU runs. Full ranking

RTX 509080RTX Pro 600067RTX 5070 Ti55M5 Ultra 256GB50RTX 409040RX 7900 XTX40RTX 309038RX 7800 XT 16GB35RTX 5090 Laptop 24GB34RX 7900 XT32

Prompt processing, same model

M5 Ultra 256GB1800RTX 40901949RX 7900 XTX640RTX 3090561RX 7900 XT800M3 Ultra 512GB372V100 32GB941Radeon AI PRO R9700 32GB1287RTX 3080 20GB997Arc Pro B70512

Reports by model size

Qwen3.8 27B819Qwen3.8 125B · 6B active676Qwen3.6 35B · 3B active242DeepSeek V4 Flash 284B · 13B active133DeepSeek V4.1 Flash 552B · 16B active92Qwen3.6 27B81Gemma 4 26B · 4B active61GLM-5.3 320B · 18B active57

Quants

Q4_K_M395UD-Q4_K_XL189NVFP4174Q8_01194-bit106IQ4_XS86IQ3_XXS85IQ3_S84Q4_K_XL83Q480