readygpu

Mellum2.1 12B A2.5B Thinking

12.1B total, 2.5B active · 128k context · Apache 2.0 · released 2026-09-20

chatcodingagenticreasoning

In short

Mellum2.1 12B A2.5B Thinking is an open model from JetBrains with 12.1B parameters (2.5B active per token). At Q6_K, Mellum2.1 12B A2.5B Thinking needs about 12 GB of memory at 8k context and 12 GB at 8k, counting the weights, the KV cache and runtime buffers. The cheapest hardware in Canada that runs Mellum2.1 12B A2.5B Thinking fully in memory at 8k context is the Intel Arc A770 16GB, $655 at Amazon, at about 55 tok/s (estimated).

Updated . Speeds are estimates; how they are worked out.

What to run it on in Canada

Fully in memory at 8k context, priced at retailers today.

All hardware that runs it

Everything we price, best value first. Two cards split the model by layers.

HardwareQuantFitsDecodePricetok/s per 1,000
Intel Arc A770 16GBQ8_0fits55 tok/s$65583.3
AMD Radeon RX 9060 XT 16GBQ8_0fits37 tok/s$80046.8
AMD Radeon RX 9070 XTQ8_0fits75 tok/s$1,05071.3
Intel Arc Pro B60Q8_0fits45 tok/s$1,06042.0
NVIDIA GeForce RTX 5060 Ti 16GBQ8_0fits59 tok/s$1,12052.8
AMD Radeon RX 7800 XTQ8_0fits73 tok/s$1,39752.3
AMD Radeon RX 7600 XTQ8_0fits34 tok/s$1,44023.4
NVIDIA GeForce RTX 5070 TiQ8_0fits118 tok/s$1,69069.9
NVIDIA GeForce RTX 4070 Ti SuperQ8_0fits89 tok/s$1,79049.5
NVIDIA GeForce RTX 4080 SuperQ8_0fits97 tok/s$2,00048.6
AMD Radeon RX 7900 XTQ8_0fits94 tok/s$2,08344.9
NVIDIA GeForce RTX 5080Q8_0fits127 tok/s$2,37053.4
NVIDIA GeForce RTX 3090Q8_0fits123 tok/s$2,80044.1
AMD Radeon AI PRO R9700Q8_0fits75 tok/s$2,80026.8
NVIDIA RTX PRO 6000 BlackwellQ8_0fits236 tok/s$3,13375.4
NVIDIA GeForce RTX 3090 TiQ8_0fits133 tok/s$3,99933.2
GMKtec EVO-X2 Ryzen AI Max+ 395 128GBQ8_0fits30 tok/s$5,0995.9
NVIDIA GeForce RTX 4090Q8_0fits133 tok/s$7,10718.7
ASUS Ascent GX10 128GBQ8_0fits36 tok/s$8,2994.3
NVIDIA GeForce RTX 5090Q8_0fits236 tok/s$10,30022.9
Mac Studio M4 Max 128GBQ8_0fits76 tok/s––
Mac mini M4 Pro 64GBQ8_0fits38 tok/s––
AMD Radeon RX 7900 XTXQ8_0fits112 tok/s––
Framework Desktop Ryzen AI Max+ 395 128GBQ8_0fits30 tok/s––
Mac Studio M3 Ultra 512GBQ8_0fits113 tok/s––
NVIDIA DGX Spark 128GBQ8_0fits36 tok/s––
Mac Studio M3 Ultra 256GBQ8_0fits113 tok/s––
Mac mini M4 Pro 48GBQ8_0fits38 tok/s––
NVIDIA GeForce RTX 4060 Ti 16GBQ8_0fits38 tok/s––
Mac Studio M4 Max 64GBQ8_0fits76 tok/s––
Mac Studio M3 Ultra 96GBQ8_0fits113 tok/s––
Framework Desktop Ryzen AI Max+ 395 64GBQ8_0fits30 tok/s––
Mac mini M4 32GBQ8_0fits17 tok/s––
NVIDIA Tesla P40Q8_0fits46 tok/s––
NVIDIA GeForce RTX 3060 12GBQ6_Koffload51 tok/s$497102.2
Intel Arc B580Q6_Koffload44 tok/s$56078.0
NVIDIA GeForce RTX 3080 10GBQ6_Koffload44 tok/s$76957.2
NVIDIA GeForce RTX 5070Q6_Koffload90 tok/s$1,18176.2
NVIDIA GeForce RTX 4070 SuperQ6_Koffload69 tok/s$1,40049.6

1 other card or machine we price cannot hold it at this context.

Memory needed

Weights, KV cache and runtime buffers at each context length (KV cache at fp16).

QuantFile@8k@32k@64k@128k
Q6_Kgguf10.7 GB11.6 GB12.0 GB12.7 GB13.9 GB
Q8_0gguf12.0 GB12.9 GB13.4 GB14.0 GB15.3 GB

Run it

Copy a command for the runtime you use.

  • Q6_K
    • Ollamaollama run hf.co/bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q6_K
    • LM Studiolms get bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF@q6_k
    • llama.cppllama-server -hf bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q6_K
  • Q8_0
    • Ollamaollama run hf.co/bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q8_0
    • LM Studiolms get bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF@q8_0
    • llama.cppllama-server -hf bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q8_0

Notes

Thinking model; serve with vLLM using the qwen3 reasoning parser (and hermes tool-call parser for tools). GGUF builds and the MTP head are not yet released.

Questions

How much VRAM does Mellum2.1 12B A2.5B Thinking need?
At Q6_K, Mellum2.1 12B A2.5B Thinking needs about 12 GB of memory at 8k context and 12 GB at 8k, counting the weights, the KV cache and runtime buffers.
What is the best GPU for running Mellum2.1 12B A2.5B Thinking locally?
The cheapest hardware in Canada that runs Mellum2.1 12B A2.5B Thinking fully in memory at 8k context is the Intel Arc A770 16GB, $655 at Amazon, at about 55 tok/s (estimated). The fastest we price is the NVIDIA RTX PRO 6000 Blackwell, at about 236 tok/s.
Can I run Mellum2.1 12B A2.5B Thinking on a 16 GB GPU?
Yes: at Q8_0 it fits fully in 16 GB at 8k context. On the NVIDIA GeForce RTX 5080 it runs at about 127 tok/s (estimated).
Can I run Mellum2.1 12B A2.5B Thinking on a 24 GB GPU?
Yes: at Q8_0 it fits fully in 24 GB at 8k context. On the NVIDIA GeForce RTX 3090 Ti it runs at about 133 tok/s (estimated).
How do I run Mellum2.1 12B A2.5B Thinking locally?
With Ollama: ollama run hf.co/bartowski/Mellum2.1-12B-A2.5B-Thinking-GGUF:Q6_K (the Q6_K build). LM Studio and llama.cpp commands are on this page.
What licence is Mellum2.1 12B A2.5B Thinking released under?
Apache 2.0, which allows commercial use.