[whatmodel.sh](../../README.md) / InternLM2.5 20B Chat
# InternLM2.5 20B Chat
Shanghai AI Laboratory (InternLM) · released 2024-08 · 19.9B params · other
**score 47.0** ±1.5 · #437 of 732 · from 7 benchmark results across 4 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks InternLM2.5 20B Chat among every model on that task or test, worst to best.
task vs the others rank counts tests
coding ▁▁▁▃▅▇●▇▅▄▃▁ #237 of 469 19% 4
chat & writing ▁▁▁▃▆●▆█▆▇▇▃ #316 of 402* 74% 2
coding · #237 of 469 · 19% of its score
Beats [Qwen3.5 4B](qwen3-5-4b.md), [Qwen3-VL 32B Instruct](qwen3-vl-32b-instruct.md), [Solar Open 100B (102B-A12B)](solar-open-100b.md) here, though they rank higher overall. Trails [QwQ-32B-Preview](qwq-32b-preview.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
CodeReviewQA: solution identifica… 70.7% ▃··▁▂▂▃▂█●▄▂ beats 38/47 medium · avg 51% 6.9% sharp · at its level · noisy
CodeReviewQA: change localisation… 55.3% █▂▁▁▂▂▂▃▅▅●▇ beats 38/47 hard · avg 33% 6.5% sharp · noisy
CodeReviewQA: automated code refi… 33.4% ▁▁▁▁▁▁▂▂▃●▅▄ beats 32/47 hard · avg 32% 5.4% at its level
CodeReviewQA: change type recogni… 75.7% ▁▁▁······▁▃● beats 30/47 easy · avg 67% 0.2% blunt · at its level · noisy
chat & writing · #316 of 402 · 74% of its score
Beats [Gemma 4 E4B](gemma-4-e4b.md), [OpenThinker3 7B](openthinker3-7b.md), [Code Llama 70B Instruct](codellama-70b-instruct.md) here, though they rank higher overall. Trails [LFM2 350M](lfm2-350m.md), [Gemma 3 1B](gemma-3-1b.md), [Qwen3 0.6B](qwen3-0-6b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
Arena-Hard (v0.1, GPT-4-1106 judg… 31.2% ▂▁▁▅█●▄▃▃▅▃▁ beats 26/60 medium · avg 47% 52% sharp · at its level
LMArena (text) 1190.8 ▁▁▁▂●▃▃▆▆▇█▃ beats 64/385 avg 1260 23% blunt
other tests (1)
test result vs the others in its run difficulty counts why
Open PT-LLM Leaderboard average (… 56.9% ▃▁▁▁▁▁▁▁●█▅▁ beats 81/276 medium · avg 72% 5.5% at its level · noisy
source results share of the score
Arena-Hard-Auto v0.1 leaderboard 1 52% ▇▇▇▇▇▇
LMArena leaderboard dataset 1 23% ▇▇▇
CodeReviewQA: The Code Review Comprehen… 4 19% ▇▇
Open Portuguese LLM Leaderboard 1 6% ▇
prior: pull toward the middle 1% ▇
weighted average = 47.0
Licences: Arena-Hard-Auto v0.1 leaderboard ([Apache-2.0](../../licenses/arena-hard-auto-v0-1-Apache-2.0.txt)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · CodeReviewQA: The Code Review Comprehension Assessment for Large Language Models ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt))
## How to run it
### llama.cpp (recommended)
1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md)
```sh
# macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
# Windows
irm https://llama.app/install.ps1 | iex
```
2. Run InternLM2.5 20B Chat (Q4_K_M, needs ~12 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/internlm2_5-20b-chat-GGUF/tree/b33fe191316ba32944057d46cd15c87d821c9495))
```sh
llama cli -hf lmstudio-community/internlm2_5-20b-chat-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf lmstudio-community/internlm2_5-20b-chat-GGUF:Q4_K_M
```
Other versions: Q8_0 (~21.1 GB) `llama cli -hf lmstudio-community/internlm2_5-20b-chat-GGUF:Q8_0`
### LM Studio
1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download)
2. Run InternLM2.5 20B Chat (Q4_K_M, needs ~12 GB, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/internlm2_5-20b-chat-GGUF/tree/b33fe191316ba32944057d46cd15c87d821c9495))
```sh
lms get https://huggingface.co/lmstudio-community/internlm2_5-20b-chat-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q8_0 (~21.1 GB) `lms get https://huggingface.co/lmstudio-community/internlm2_5-20b-chat-GGUF@Q8_0 -y` · 8bit (~21.1 GB, Mac only) `lms get https://huggingface.co/mlx-community/internlm2_5-20b-chat_8bit -y`
### Ollama
1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download)
```sh
# Linux
curl -fsSL https://ollama.com/install.sh | sh
```
2. Run InternLM2.5 20B Chat (Q4_K_M, needs ~12.44 GB, from [ollama.com](https://ollama.com/library/internlm2/tags))
```sh
ollama run internlm2:20b-chat-v2.5-q4_K_M
```
Other versions: Q8_0 (~24.38 GB) `ollama run internlm2:20b-chat-v2.5-q8_0`
### MLX (Mac)
1. Install MLX (Mac). Apple silicon Macs only, text only.
```sh
# with uv
uv tool install mlx-lm
# or
pip install mlx-lm
```
2. Run InternLM2.5 20B Chat (8bit, needs ~21.1 GB, Mac only, from [huggingface.co/mlx-community](https://huggingface.co/mlx-community/internlm2_5-20b-chat_8bit/tree/4095132a0deb21cfd5f0342e81ad605b06dbc91a))
```sh
mlx_lm.chat --model mlx-community/internlm2_5-20b-chat_8bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model mlx-community/internlm2_5-20b-chat_8bit
```
Model repository and documentation: [huggingface.co/internlm](https://huggingface.co/internlm/internlm2_5-20b-chat)
[interactive version](https://whatmodel.sh/model/internlm2-5-20b-chat) · [JSON](https://whatmodel.sh/api?format=json&all=1)