Skip to content

InternLM 2.5 7B

by Shanghai AI Lab · internlm-2.5 family

7B

parameters

text-generation code-generation reasoning multilingual tool-use math

InternLM 2.5 7B is Shanghai AI Lab's versatile open-source language model featuring strong reasoning, multilingual support, and tool-use capabilities. With a massive 1M token context window, it handles extremely long documents and complex multi-step tasks that most models of its size cannot. The model excels at mathematical reasoning, code generation, and function calling, making it a strong choice for agentic workflows. Its Apache 2.0 license and competitive benchmark scores position it as an excellent option for local deployment on consumer hardware.

Quick Start with Ollama

ollama run 7b-q4_K_M
Resources Ollama Hugging Face Official Page
Creator Shanghai AI Lab
Parameters 7B
Architecture transformer-decoder
Context 1024K tokens
Released Jul 3, 2024
License Apache 2.0
Ollama internlm2:7b

Quantization Options

Format File Size VRAM Required Quality Ollama Tag
Q4_K_M rec 4.4 GB 5.5 GB 7b-q4_K_M
Q8_0 7.4 GB 9 GB 7b-q8_0
F16 14 GB 16 GB 7b-f16

Compatible Hardware

Q4_K_M requires 5.5 GB VRAM

Compatible Hardware

HardwareVRAMTypeFitEst. Speed
Mac Studio M4 Ultra 512GB512 GBmacRuns~149 tok/s
Mac Pro M2 Ultra 192GB192 GBmacRuns~145 tok/s
Mac Studio M4 Ultra 192GB192 GBmacRuns~149 tok/s
Mac Studio M4 Max 128GB128 GBmacRuns~99 tok/s
MacBook Pro M4 Max 128GB128 GBmacRuns~99 tok/s
MacBook Pro M5 Max 128GB128 GBmacRuns~99 tok/s
NVIDIA RTX PRO 6000 Blackwell96 GBgpuRuns~349 tok/s
MacBook Pro M3 Max 96GB96 GBmacRuns~73 tok/s
Mac mini M4 Pro 64GB64 GBmacRuns~50 tok/s
Mac Studio M4 Max 64GB64 GBmacRuns~99 tok/s
MacBook Pro M4 Max 64GB64 GBmacRuns~99 tok/s
MacBook Pro M5 Max 64GB64 GBmacRuns~99 tok/s
NVIDIA RTX 6000 Ada Generation48 GBgpuRuns~175 tok/s
NVIDIA RTX A600048 GBgpuRuns~140 tok/s
NVIDIA RTX PRO 5000 Blackwell48 GBgpuRuns~175 tok/s
Mac mini M4 Pro 48GB48 GBmacRuns~50 tok/s
MacBook Pro M3 Max 48GB48 GBmacRuns~73 tok/s
MacBook Pro M4 Max 48GB48 GBmacRuns~99 tok/s
MacBook Pro M4 Pro 48GB48 GBmacRuns~50 tok/s
MacBook Pro M5 Max 48GB48 GBmacRuns~75 tok/s
MacBook Pro M5 Pro 48GB48 GBmacRuns~50 tok/s
Mac Studio M4 Max 36GB36 GBmacRuns~99 tok/s
MacBook Pro M3 Pro 36GB36 GBmacRuns~27 tok/s
MacBook Pro M5 Max 36GB36 GBmacRuns~75 tok/s
NVIDIA RTX 5000 Ada Generation32 GBgpuRuns~131 tok/s
NVIDIA GeForce RTX 509032 GBgpuRuns~326 tok/s
iMac M4 32GB32 GBmacRuns~22 tok/s
Mac mini M4 32GB32 GBmacRuns~22 tok/s
MacBook Air M5 32GB32 GBmacRuns~22 tok/s
MacBook Air M4 32GB32 GBmacRuns~22 tok/s
MacBook Pro M5 32GB32 GBmacRuns~22 tok/s
AMD Radeon RX 7900 XTX24 GBgpuRuns~175 tok/s
NVIDIA GeForce RTX 3090 Ti24 GBgpuRuns~183 tok/s
NVIDIA GeForce RTX 309024 GBgpuRuns~170 tok/s
NVIDIA GeForce RTX 409024 GBgpuRuns~183 tok/s
NVIDIA RTX A500024 GBgpuRuns~140 tok/s
iMac M3 24GB24 GBmacRuns~18 tok/s
Mac mini M2 24GB24 GBmacRuns~18 tok/s
Mac mini M4 Pro 24GB24 GBmacRuns~50 tok/s
MacBook Air M2 24GB24 GBmacRuns~18 tok/s
MacBook Air M4 24GB24 GBmacRuns~22 tok/s
MacBook Air M5 24GB24 GBmacRuns~22 tok/s
MacBook Pro M4 Pro 24GB24 GBmacRuns~50 tok/s
MacBook Pro M5 24GB24 GBmacRuns~22 tok/s
MacBook Pro M5 Pro 24GB24 GBmacRuns~50 tok/s
AMD Radeon RX 7900 XT20 GBgpuRuns~145 tok/s
NVIDIA RTX 4000 Ada Generation20 GBgpuRuns~65 tok/s
MacBook Pro M3 Pro 18GB18 GBmacRuns~27 tok/s
AMD Radeon RX 6800 XT16 GBgpuRuns~93 tok/s
AMD Radeon RX 6900 XT16 GBgpuRuns~93 tok/s
AMD Radeon RX 7800 XT16 GBgpuRuns~113 tok/s
AMD Radeon RX 9060 XT 16GB16 GBgpuRuns~98 tok/s
AMD Radeon RX 9070 XT16 GBgpuRuns~118 tok/s
AMD Radeon RX 907016 GBgpuRuns~98 tok/s
Intel Arc A77016 GBgpuRuns~102 tok/s
NVIDIA GeForce RTX 4060 Ti 16GB16 GBgpuRuns~52 tok/s
NVIDIA GeForce RTX 4070 Ti Super16 GBgpuRuns~122 tok/s
NVIDIA GeForce RTX 4080 Super16 GBgpuRuns~134 tok/s
NVIDIA GeForce RTX 408016 GBgpuRuns~130 tok/s
NVIDIA GeForce RTX 5060 Ti 16GB16 GBgpuRuns~81 tok/s
NVIDIA GeForce RTX 5070 Ti16 GBgpuRuns~163 tok/s
NVIDIA GeForce RTX 508016 GBgpuRuns~175 tok/s
NVIDIA RTX A400016 GBgpuRuns~81 tok/s
iMac M1 16GB16 GBmacRuns~12 tok/s
iMac M4 16GB16 GBmacRuns~22 tok/s
Mac mini M1 16GB16 GBmacRuns~12 tok/s
Mac mini M4 16GB16 GBmacRuns~22 tok/s
MacBook Air M2 16GB16 GBmacRuns~18 tok/s
MacBook Air M4 16GB16 GBmacRuns~22 tok/s
MacBook Air M3 16GB16 GBmacRuns~18 tok/s
MacBook Air M5 16GB16 GBmacRuns~22 tok/s
MacBook Pro M2 Pro 16GB16 GBmacRuns~36 tok/s
MacBook Pro M1 16GB16 GBmacRuns~12 tok/s
MacBook Pro M5 16GB16 GBmacRuns~22 tok/s
AMD Radeon RX 6700 XT12 GBgpuRuns~70 tok/s
AMD Radeon RX 7700 XT12 GBgpuRuns~79 tok/s
Intel Arc B58012 GBgpuRuns~83 tok/s
NVIDIA GeForce RTX 3060 12GB12 GBgpuRuns~65 tok/s
NVIDIA GeForce RTX 3080 12GB12 GBgpuRuns~166 tok/s
NVIDIA GeForce RTX 4070 Super12 GBgpuRuns~92 tok/s
NVIDIA GeForce RTX 4070 Ti12 GBgpuRuns~92 tok/s
NVIDIA GeForce RTX 407012 GBgpuRuns~92 tok/s
NVIDIA GeForce RTX 507012 GBgpuRuns~122 tok/s
NVIDIA GeForce GTX 1080 Ti11 GBgpuRuns~88 tok/s
NVIDIA GeForce RTX 2080 Ti11 GBgpuRuns~112 tok/s
Intel Arc B57010 GBgpuRuns~69 tok/s
NVIDIA GeForce RTX 3080 10GB10 GBgpuRuns~138 tok/s
AMD Radeon RX 6600 XT8 GBgpuRuns~47 tok/s
AMD Radeon RX 76008 GBgpuRuns~52 tok/s
AMD Radeon RX 9060 XT 8GB8 GBgpuRuns~49 tok/s
Intel Arc A7508 GBgpuRuns~93 tok/s
NVIDIA GeForce GTX 10708 GBgpuRuns~47 tok/s
NVIDIA GeForce RTX 2060 Super8 GBgpuRuns~81 tok/s
NVIDIA GeForce RTX 2070 Super8 GBgpuRuns~81 tok/s
NVIDIA GeForce RTX 2080 Super8 GBgpuRuns~90 tok/s
NVIDIA GeForce RTX 30508 GBgpuRuns~41 tok/s
NVIDIA GeForce RTX 3060 Ti8 GBgpuRuns~81 tok/s
NVIDIA GeForce RTX 30708 GBgpuRuns~81 tok/s
NVIDIA GeForce RTX 4060 Ti 8GB8 GBgpuRuns~52 tok/s
NVIDIA GeForce RTX 40608 GBgpuRuns~49 tok/s
NVIDIA GeForce RTX 50508 GBgpuRuns~41 tok/s
NVIDIA GeForce RTX 5060 Ti 8GB8 GBgpuRuns~81 tok/s
NVIDIA GeForce RTX 50608 GBgpuRuns~61 tok/s
MacBook Air M1 8GB8 GBmacRuns~12 tok/s
MacBook Air M2 8GB8 GBmacRuns~18 tok/s
NVIDIA GeForce GTX 1660 Super6 GBgpuRuns (tight)~61 tok/s
NVIDIA GeForce RTX 20606 GBgpuRuns (tight)~61 tok/s

Benchmark Scores

72.0
mmlu