[whatmodel.sh](../../README.md) / Seed-OSS 36B Instruct
# Seed-OSS 36B Instruct
ByteDance Seed · released 2025-08 · 36.2B params · apache-2.0
**score 62.1** ±1.0 · #180 of 732 · from 21 benchmark results across 2 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks Seed-OSS 36B Instruct among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃▆█▄●▅▄▄▂▁ #148 of 313 15% 4
coding ▁▁▁▃▅▇█▇●▄▃▁ #131 of 469 11% 3
chat & writing ▁▁▁▃▆▄▆●▆▇▇▃ #196 of 402* 1.6% 1
math ▁▁▃▅█▆▄●▄▂▁▁ #147 of 431 19% 3
agent tasks · #148 of 313 · 15% of its score
Beats [DeepSeek V3.2 Exp](deepseek-v3-2-exp.md), [o1 (2024-12-17)](o1-2024-12-17.md), [o4-mini](o4-mini.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Large](jamba-1-5-large.md), [GPT-4o (2024-05-13)](gpt-4o-2024-05-13.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
SWE-bench Verified (maker) 56.0% ▁▁▁▁▁▁▂●█▆▁▁ beats 46/138 hard · avg 16% 5.1% sharp · at its level · small run
τ-bench Retail (maker) 70.4% ▁····▃·▂▇█●▆ beats 21/28 hard · avg 33% 3.9% at its level · small run
τ-bench Airline (maker) 46.0% ▁▁▁·▅▄▁●▇▁▁▇ beats 16/28 hard · avg 25% 3.8% at its level · small run
Multi-SWE-bench (scaffold per not… 17.0% █··█·······● beats 2/2 very hard · avg 5.9% 2.1% small run
coding · #131 of 469 · 11% of its score
Beats [Qwen3.5 Plus](qwen3-5-plus.md), [Qwen3-Coder-Next (80B-A3B)](qwen3-coder-next.md), [Gemini 2.5 Pro](gemini-2-5-pro.md) here, though they rank higher overall. Trails [Llama 3.1 Nemotron Ultra 253B v1](llama-3-1-nemotron-ultra-253b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
SWE-bench Verified (maker) 56.0% ▁▁▁▁▁▁▂●█▆▁▁ beats 46/138 hard · avg 16% 5.1% sharp · at its level · small run
LiveCodeBench v6 (Feb–May 2025) (… 67.4% ▁▁▃▁█▅▄●▅▅▃█ beats 18/38 medium · avg 41% 4.2% at its level
Multi-SWE-bench (scaffold per not… 17.0% █··█·······● beats 2/2 very hard · avg 5.9% 2.1% small run
chat & writing · #196 of 402 · 1.6% of its score
Trails [LFM2.5 1.2B Instruct](lfm2-5-1-2b-instruct.md), [LFM2.5 1.2B Thinking](lfm2-5-1-2b-thinking.md), [LFM2.5 8B-A1B](lfm2-5-8b-a1b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
IFEval (aggregation unspecified) … 85.8% ▁▂▄▂▂▄▆●█▅▂▁ beats 80/145 easy · avg 80% 1.6% typical
math · #147 of 431 · 19% of its score
Beats [Claude Sonnet 4.6](claude-sonnet-4-6.md), [MiMo-V2.5 (310B-A15B)](mimo-v2-5.md), [MiniMax-M2.7](minimax-m2-7.md) here, though they rank higher overall. Trails [Olmo 3 32B Think](olmo-3-32b-think.md), [Nemotron-Cascade-8B](nemotron-cascade-8b.md), [Olmo 3.1 32B Think](olmo-3-1-32b-think.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
OTIS mock AIME 2024–2025 67.5% ▃▃▄▃▃▃●▇█▃▂▅ beats 97/197 very hard · avg 7.2% 7.7% sharp · at its level · noisy
AIME 2024 (maker) 91.7% ▁▁▁▂▁▃▁▅██●▃ beats 78/85 easy · avg 69% 5.8% sharp
AIME 2025 (maker) 84.7% ▁·▁▁▁▂▅▆█●▅▁ beats 98/133 medium · avg 56% 5.5% sharp
other tests (11)
test result vs the others in its run difficulty counts why
GPQA Diamond 71.5% ▁▁▁▁▂▄▆▄●▇█▅ beats 108/211 hard · avg 39% 27% sharp · at its level · several runs
MMLU (maker) 87.4% ▁▁▁▁▁▂▂▅██▆● beats 201/214 medium · avg 70% 6.0% typical
Humanity's Last Exam (maker) 10.1% ▁▁▁▄●▆▇██▆▁▃ beats 19/111 very hard · avg 4.0% 5.4% typical
MMLU-Pro (maker) 82.7% ▁▁·▁▁▁▃▂▄▅●▁ beats 151/188 medium · avg 63% 4.7% typical
MMMLU (maker) 78.4% ▂▁▁▂▃▂▄▄▅●█▃ beats 26/46 medium · avg 64% 4.0% typical
SuperGPQA (maker) 55.7% ▂▁▂▂▅▆▃▇●▇█▃ beats 27/46 medium · avg 37% 3.2% at its level
SimpleQA (maker) 9.7% ▄▄▄▆●▆███·▂█ beats 10/32 very hard · avg 6.1% 2.7% typical
KOR-Bench (maker) 70.6% ▄·······●··▄ beats 2/3 medium · avg 55% 2.4% at its level
Chess Puzzles (Epoch) 13.0% █▁▁▁▁●▃▃▃▂▁▁ beats 64/137 very hard · avg 5.8% 2.3% sharp · too hard for it · noisy
RULER 128k (maker) 94.6% ▃··▅▄█▅█▇·▃● beats 33/33 easy · avg 72% 1.5% typical
BeyondAIME (maker) 65.0% ▄·▂·▂▄··▄▄●█ beats 10/15 medium · avg 37% 1.0% at its level · noisy
source results share of the score
ByteDance-Seed/Seed-OSS-36B-Instruct mo… 18 73% ▇▇▇▇▇▇▇▇▇
Epoch AI 3 26% ▇▇▇
prior: pull toward the middle 0% ▇
weighted average = 62.1
Licences: ByteDance-Seed/Seed-OSS-36B-Instruct model card ([apache-2.0](../../licenses/ByteDance-Seed--Seed-OSS-36B-Instruct.txt)) · Epoch AI ([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 Seed-OSS 36B Instruct (gguf-q4_k_m, needs ~22.22 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/Seed-OSS-36B-Instruct-GGUF/tree/main))
```sh
llama cli -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
```
Other versions: gguf-q8_0 (~43.94 GB) `llama cli -hf unsloth/Seed-OSS-36B-Instruct-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 Seed-OSS 36B Instruct (gguf-q4_k_m, needs ~22.22 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/Seed-OSS-36B-Instruct-GGUF/tree/main))
```sh
lms get https://huggingface.co/unsloth/Seed-OSS-36B-Instruct-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: gguf-q8_0 (~43.94 GB) `lms get https://huggingface.co/unsloth/Seed-OSS-36B-Instruct-GGUF@Q8_0 -y` · 4bit (~20.3 GB, Mac only) `lms get https://huggingface.co/lmstudio-community/Seed-OSS-36B-Instruct-MLX-4bit -y` · 8bit (~38.4 GB, Mac only) `lms get https://huggingface.co/lmstudio-community/Seed-OSS-36B-Instruct-MLX-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 Seed-OSS 36B Instruct (gguf-q4_k_m, needs ~22.22 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/Seed-OSS-36B-Instruct-GGUF/tree/main))
```sh
ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_M
```
Other versions: gguf-q8_0 (~43.94 GB) `ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-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 Seed-OSS 36B Instruct (4bit, needs ~20.3 GB, Mac only, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/Seed-OSS-36B-Instruct-MLX-4bit/tree/aedfa6b5de1525fa96e3f12d81ba00f2cbfc22a5))
```sh
mlx_lm.chat --model lmstudio-community/Seed-OSS-36B-Instruct-MLX-4bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model lmstudio-community/Seed-OSS-36B-Instruct-MLX-4bit
```
Other versions: 8bit (~38.4 GB, Mac only) `mlx_lm.chat --model lmstudio-community/Seed-OSS-36B-Instruct-MLX-8bit`
Model repository and documentation: [huggingface.co/ByteDance-Seed](https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct)
[interactive version](https://whatmodel.sh/model/seed-oss-36b) · [JSON](https://whatmodel.sh/api?format=json&all=1)