[whatmodel.sh](../../README.md) / Nemotron-Cascade-8B # Nemotron-Cascade-8B NVIDIA · released 2025-12 · 8.19B params · nvidia-open-model-license **score 58.0** ±1.0 · #227 of 732 · from 12 benchmark results across 1 source · † no independently run results verified Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at _{● marks Nemotron-Cascade-8B among every model on that task or test, worst to best.}

task            vs the others          rank  counts  tests
agent tasks     ▁▁▃▆●▄▅▅▄▄▂▁   #207 of 313*    7.6%      2
coding          ▁▁▁▃▅▇█●▅▄▃▁    #202 of 469     50%      4
chat & writing  ▁▁▁▃▆▄▆█▆●▇▃   #112 of 402*    1.9%      1
math            ▁▁▃▅█▆▄●▄▂▁▁   #138 of 431*     19%      2

agent tasks · #207 of 313 · 7.6% of its score

Beats [o1 (2024-12-17)](o1-2024-12-17.md), [o3-mini](o3-mini.md), [DeepSeek R1 (Jan '25)](deepseek-r1.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Mini](jamba-1-5-mini.md), [Mellum2 12B-A2.5B Instruct](mellum2-12b-a2-5b-instruct.md) here, though they rank lower overall.

test                        result  vs the others in its run    difficulty        counts  why
SWE-bench Verified (maker)   37.2%  ▁▁▁▁▁▁●▄█▆▁▁ beats 26/138   hard · avg 19%      6.1%  sharp · at its level · small run
BFCL v3 (maker)              64.4%  ▁▁▁▁▁▃▅█●▅▄▁ beats 122/171  medium · avg 58%    1.5%  blunt · at its level · small run

coding · #202 of 469 · 50% of its score

Beats [Gemini 2.5 Flash](gemini-2-5-flash.md), [Qwen3 30B-A3B Instruct 2507](qwen3-30b-a3b-instruct-2507.md), [Qwen3.5 4B](qwen3-5-4b.md) here, though they rank higher overall. Trails [QwQ-32B-Preview](qwq-32b-preview.md), [Ministral 3 3B Reasoning 2512](ministral-3-3b-reasoning.md), [DeepSeek-Coder-V2 Instruct (0724)](deepseek-coder-v2-0724.md) here, though they rank lower overall.

test                                result  vs the others in its run   difficulty            counts  why
LiveCodeBench Pro 2025Q2 (Medium)…    6.4%  ▃▃····▃·●·▃▃ beats 5/7     very hard · avg 2.6%     26%  sharp · too hard for it
LiveCodeBench Pro 2025Q2 (Easy) (…   65.7%  ▄·▄····▄▄▄●▄ beats 5/7     medium · avg 46%         12%  at its level
SWE-bench Verified (maker)           37.2%  ▁▁▁▁▁▁●▄█▆▁▁ beats 26/138  hard · avg 19%          6.1%  sharp · at its level · small run
LiveCodeBench v6 (Aug 2024–May 20…   71.1%  ▄▄···▄·▄▄●·▄ beats 5/7     medium · avg 58%        5.4%  at its level

chat & writing · #112 of 402 · 1.9% of its score

Beats [LongCat 2.0 (1.6T-A48B)](longcat-2-0.md), [GPT-5.4 nano](gpt-5-4-nano.md), [Step 3.5 Flash (196B-A11B)](step-3-5-flash.md) here, though they rank higher overall.

test                                result  vs the others in its run  difficulty      counts  why
IFEval prompt-level strict (maker)   90.2%  ▁▁··▂··▂▇█▃● beats 24/26  easy · avg 77%    1.9%  small run

math · #138 of 431 · 19% of its score

Beats [Claude Sonnet 4.6](claude-sonnet-4-6.md), [MiniMax-M3](minimax-m3.md), [MiMo-V2.5 (310B-A15B)](mimo-v2-5.md) here, though they rank higher overall. Trails [Olmo 3 32B Think](olmo-3-32b-think.md) here, though they rank lower overall.

test               result  vs the others in its run   difficulty        counts  why
AIME 2025 (maker)   80.1%  ▁·▁▁▁▂▅▆█●▅▁ beats 85/133  medium · avg 56%     10%  sharp · small run
AIME 2024 (maker)   89.5%  ▁▁▁▂▁▃▁▅██●▃ beats 70/85   easy · avg 70%      8.6%  sharp · small run

other tests (4)

test                                result  vs the others in its run    difficulty        counts  why
GPQA Diamond (maker)                 66.5%  ▁▁▁▁▂▄▅▅●██▄ beats 141/322  hard · avg 46%       12%  sharp · at its level · small run
MMLU (maker)                         83.7%  ▁▁▁▁▁▂▂▅██●▄ beats 181/214  medium · avg 71%    6.7%  small run
MMLU-Pro (maker)                     75.7%  ▁▁·▁▁▁▃▂▄●█▁ beats 104/188  medium · avg 60%    5.2%  at its level · small run
IFBench (aggregation unspecified)…   40.8%  ▂▆▂▄●▄▂▄▄▆█▇ beats 27/86    hard · avg 30%      3.4%  at its level · small run

## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md)

source                               results  share of the score
NVIDIA (submitted / maker-reported)       12  100% ▇▇▇▇▇▇▇▇▇▇▇▇
prior: pull toward the middle                   0% ▇
weighted average                              = 58.0

_{Licences: NVIDIA ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/))} ## 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 Nemotron-Cascade-8B (Q4_K_M, needs ~5.03 GB, from [huggingface.co/bartowski](https://huggingface.co/bartowski/nvidia_Nemotron-Cascade-8B-GGUF/tree/a276f0387d3d2708030e130a4c9e2b3e291a7616)) ```sh llama cli -hf bartowski/nvidia_Nemotron-Cascade-8B-GGUF:Q4_K_M ``` 3. Or start a local server and open the address it prints (http://localhost:9931) ```sh llama serve -hf bartowski/nvidia_Nemotron-Cascade-8B-GGUF:Q4_K_M ``` Other versions: Q8_0 (~8.71 GB) `llama cli -hf bartowski/nvidia_Nemotron-Cascade-8B-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 Nemotron-Cascade-8B (Q4_K_M, needs ~5.03 GB, from [huggingface.co/bartowski](https://huggingface.co/bartowski/nvidia_Nemotron-Cascade-8B-GGUF/tree/a276f0387d3d2708030e130a4c9e2b3e291a7616)) ```sh lms get https://huggingface.co/bartowski/nvidia_Nemotron-Cascade-8B-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 (~8.71 GB) `lms get https://huggingface.co/bartowski/nvidia_Nemotron-Cascade-8B-GGUF@Q8_0 -y` · 4bit (~4.6 GB, Mac only) `lms get https://huggingface.co/introvoyz041/Nemotron-Cascade-8B-mlx-4Bit -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 Nemotron-Cascade-8B (Q4_K_M, needs ~5.03 GB, from [huggingface.co/bartowski](https://huggingface.co/bartowski/nvidia_Nemotron-Cascade-8B-GGUF/tree/a276f0387d3d2708030e130a4c9e2b3e291a7616)) ```sh ollama run hf.co/bartowski/nvidia_Nemotron-Cascade-8B-GGUF:Q4_K_M ``` Other versions: Q8_0 (~8.71 GB) `ollama run hf.co/bartowski/nvidia_Nemotron-Cascade-8B-GGUF: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 Nemotron-Cascade-8B (4bit, needs ~4.6 GB, Mac only, from [huggingface.co/introvoyz041](https://huggingface.co/introvoyz041/Nemotron-Cascade-8B-mlx-4Bit/tree/25a0528eec9806be8f8a195ff565e4f4aaace1fd)) ```sh mlx_lm.chat --model introvoyz041/Nemotron-Cascade-8B-mlx-4Bit ``` 3. Or start an OpenAI-compatible server at http://localhost:8080/v1 ```sh mlx_lm.server --model introvoyz041/Nemotron-Cascade-8B-mlx-4Bit ``` Model repository and documentation: [huggingface.co/nvidia](https://huggingface.co/nvidia/Nemotron-Cascade-8B) [interactive version](https://whatmodel.sh/model/nemotron-cascade-8b) · [JSON](https://whatmodel.sh/api?format=json&all=1)