BestLLMfor EN Your hardware. Your LLM. Your call.
APIOpen data Find my LLM
Model fiche

Seed-OSS 36B Instruct

By ByteDance · China

Updated 2026-07-13

chat general
Parameters
36B
License
Apache 2.0
Context
512k
VRAM (Q4)
22 GB
Released
April 2025

Overview

ByteDance's first major open release: a dense 36B model with a native 524k context — roughly 4× the competition. Apache 2.0.

When to pick this model

  • Extreme long-document analysis (codebases, books, transcripts)
  • RAG-free workflows that load everything into context
  • Dense-model deployments preferring predictable behavior
  • Apache-licensed commercial use

VRAM requirements by quantization

VRAM REQUIRED (GB)81216243248Q4_K_M22 GBQ5_K_M26 GBQ8_040 GBFP1672 GB
QuantizationVRAM required
Q4_K_M (recommended)22 GB
Q5_K_M26 GB
Q8_040 GB
FP16 (no quantization)72 GB

VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.

In practice, Seed-OSS 36B Instruct wants a 24 GB card at Q4_K_M (22 GB). Stepping up to Q8_0 nearly doubles the footprint to 40 GB, and unquantized FP16 weights take 72 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.

Without a GPU, Seed-OSS 36B Instruct needs roughly 36 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 3 tokens/sec on entry-level GPUs, on the order of 12 tokens/sec on a mid-range card, and up to 28 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.

What hardware do you need

The table below matches Seed-OSS 36B Instruct to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.

GPU memoryExample cardsBest fit for Seed-OSS 36B Instruct
8 GBRTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GBDoes not fit — needs 22 GB at Q4_K_M
12 GBRTX 5070, RTX 5070 Ti Laptop, RTX 4080 LaptopDoes not fit — needs 22 GB at Q4_K_M
16 GBRTX 5080, RTX 4080 Super, Radeon RX 9070 XTDoes not fit — needs 22 GB at Q4_K_M
24 GBRTX 4090, Radeon RX 7900 XTX, RTX 5090 LaptopQ4_K_M (22 GB used)
32 GBRTX 5090Q5_K_M (26 GB used)

Which GPU should you buy to run Seed-OSS 36B Instruct?

To run Seed-OSS 36B Instruct locally at Q4, you need ~22 GB of VRAM. The best value for this is a RTX 4090 (24 GB VRAM).

Check RTX 4090 price on Amazon →

As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.

Strengths

  • 524k native context — a record for accessible dense models
  • Dense 36B is easier to deploy than equivalent MoEs
  • Strong long-document comprehension
  • Apache 2.0

Limitations

  • Around 22 GB VRAM at Q4 (much more with full context)
  • ByteDance license terms need a careful read
  • Limited fine-tune ecosystem at launch

Typical workloads

In our catalog grid, Seed-OSS 36B Instruct is filed under Very Long Context, Book Analysis — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: everyday chat, drafting and summarization.

The 512k-token context window is large enough to hold entire codebases' worth of files or long reports in a single prompt, which is what makes local RAG and document analysis practical. The Apache 2.0 license is permissive, so shipping it inside a commercial product raises no special legal questions.

Architecture & training

Architecture: Dense · 36B · ByteDance Seed-OSS · 524k native context

Training: ByteDance — very long context (524k tokens) natively supported.

Verdict

Unmatched long-context for a dense open model — the pick when you genuinely need to load 500k+ tokens at once.

Quick start

ollama pull hf.co/ByteDance/seed-oss-36b-GGUF

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.

Similar models worth comparing

Frequently asked questions

How much VRAM does Seed-OSS 36B Instruct need?

At the recommended Q4_K_M quantization, Seed-OSS 36B Instruct needs about 22 GB of VRAM. Q8_0 takes 40 GB, and unquantized FP16 weights take 72 GB.

Can Seed-OSS 36B Instruct run without a GPU?

Yes — with roughly 36 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.

What context window does Seed-OSS 36B Instruct support?

Seed-OSS 36B Instruct supports a 512k-token context window (524,288 tokens).

Can I use Seed-OSS 36B Instruct commercially?

Yes. Seed-OSS 36B Instruct is released under Apache 2.0, a permissive open-source license that allows commercial use, modification and redistribution.

How fast is Seed-OSS 36B Instruct on consumer hardware?

Our compatibility engine estimates on the order of 12 tokens/sec on a mid-range GPU and up to 28 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.

Which quantization of Seed-OSS 36B Instruct should I download first?

Start with Q4_K_M (22 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. On a 24 GB card you can run up to Q4_K_M.

Tools

Is Seed-OSS 36B Instruct the right pick for you?

Compute self-hosted ROI → Back to catalog