gpt-oss-120b
gpt-oss-120b is an open-weights language model from OpenAI, released August 5, 2025. It has 117B parameters (mixture-of-experts, 5.8B active). The context window is 128K tokens. On the Epoch Capabilities Index it scores 140.2, rank 125 of 256 rated models.
Check it against your own PC →
Can you run gpt-oss-120b locally?
The smallest common GPU it runs on is 8 GB of VRAM (with 32 GB of RAM): IQ2_XXS, GPU + experts in RAM, roughly 31 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Too large | – | – | – | – |
| 8 GB VRAM | Runs · GPU + CPU experts | IQ2_XXS (heavy loss) | 7.2 GB | 24.1 GB | ~31 |
| 12 GB VRAM | Runs · GPU + CPU experts | IQ2_XXS (heavy loss) | 11.2 GB | 20.1 GB | ~36 |
| 16 GB VRAM | Runs · GPU + CPU experts | IQ2_M (heavy loss) | 15.2 GB | 25.4 GB | ~30 |
| 24 GB VRAM | Runs · GPU + CPU experts | IQ3_XXS (noticeable loss) | 23.2 GB | 22.7 GB | ~34 |
| 32 GB VRAM | Runs · GPU + CPU experts | IQ3_XXS (noticeable loss) | 31.2 GB | 14.7 GB | ~50 |
Ready-made quantizations on Hugging Face: ggml-org/gpt-oss-120b-GGUF · unsloth/gpt-oss-120b-GGUF · lmstudio-community/gpt-oss-120b-GGUF · HauhauCS/GPTOSS-120B-Uncensored-HauhauCS-Aggressive · bartowski/openai_gpt-oss-120b-GGUF · xCloudinfo/gpt-oss-120b-Uncensored-xCloud-GGUF
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 9.1 | SimpleBench, Mystery Game Puzzles |
| Coding | 27.7 | Aider polyglot, Terminal Bench |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| OTIS Mock AIME | 88.9% |
| GPQA Diamond | 70.5% |
| Aider Polyglot | 41.8% |
| Terminal-Bench | 18.7% |
| SimpleBench | 6.5% |
| Mystery Game Puzzles | 0.0% |
Links
Announcement · Hugging Face: openai/gpt-oss-120b · Epoch AI benchmarks
Models with a similar score
Qwen 3.6 35B-A3B · Qwen3.5-35B-A3B · Gemini 2.5 Pro (Mar 2025) · Qwen 3.5 Flash (hosted 35B-A3B) · Qwen 3.6 Flash · GLM-4.7 · Qwen3-235B-A22B-Thinking (Jul 2025) · Claude Opus 4
More from OpenAI
GPT-6.1 Terra · GPT-6.1 Sol · GPT-6 Sol · GPT-6 Luna · GPT-6 Astra · GPT-5.6 Cyber · Daybreak Red · Daybreak Blue