- Size
- 117B parametersfrom the GGUF metadata of a conversion, because this repo's own weights are packed: ggml-org/gpt-oss-120b-GGUF
- Licence
- unknown, Apache License 2.0Our reading, check the licence: You can use, change and redistribute the model, including commercially. Keep the licence and any notices, and say which files you changed; you do not have to publish your changes. It includes a patent licence that ends for anyone who sues over patents in the work.
- On the Hub since
- 4 Aug 2025the date the repo was created
- Context length
- 128k tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)GptOssForCausalLM
- On Hugging Face
- openai/gpt-oss-120b2.4M downloads in the last 30 days, 5.5k likes
Scores
- effort high75.8%±2.711 Dec 2025, log
Epoch's name: openai/gpt-oss-120b. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results
Epoch's name: gpt-oss-120b. Best closed model on this set: gpt-6.2-astra_max, 95.0% (OpenAI). All 148 results
- effort high88.9%±4.511 Dec 2025, log
Epoch's name: openai/gpt-oss-120b. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 results
- effort high; Edit format: diff41.8%Aider LLM Leaderboards, via Epoch AI
- effort high; Edit format: diff41.8%6 Aug 2025, aider.chat
Epoch's name: gpt-oss-120b, openai/gpt-oss-120b. Best closed model on this set: gpt-5-2025-08-07_high, 88.0% (OpenAI). All 72 results
- default setting4.4%via Epoch AI (error bar left out)
Epoch's name: gpt-oss-120b. Best closed model on this set: gemini-4-carbon_unknown, 84.2% (Google DeepMind). All 66 results
- effort high1.1%via Epoch AI
- effort low0.0%via Epoch AI
Epoch's name: openai/gpt-oss-120b. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results
- effort high76.3%conceptualreasoning.ai
Epoch's name: gpt-oss-120b. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results
- effort high44.4%Fiction.live leaderboard
- default setting36.1%Fiction.live leaderboard
Epoch's name: chutes/gpt-oss-120b. Best closed model on this set: gpt-5-2025-08-07_medium, 97.2% (OpenAI). All 61 results
- default setting77.1%lechmazur/writing Github repository
Epoch's name: gpt-oss-120b. Best closed model on this set: gpt-5-2025-08-07_medium, 86.0% (OpenAI). All 49 results
- effort high22.1%17.3 to 27.0conceptualreasoning.ai
Epoch's name: gpt-oss-120b. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results
- default setting56.6%METR - Measuring AI Ability to Complete Long Tasks
Epoch's name: gpt-oss-120b. Best closed model on this set: claude-mythos-preview-early, 85.2% (Anthropic). All 49 results
- default setting22.1%simple-bench.com
Epoch's name: gpt-oss-120b. Best closed model on this set: claude-fable-5-2_unknown, 94.9% (Anthropic). All 125 results
- default setting25.0%yhenon.github.io
Epoch's name: gpt-oss-120b. Best closed model on this set: claude-fable-5_high, 95.0% (Anthropic). All 29 results
- Agent: Terminus 218.7%±2.71 Nov 2025, Terminal-Bench v2 Leaderboard
- effort unknown; Agent: Terminus 218.7%±2.75 Aug 2025, tbench.ai
- Agent: Mini-SWE-Agent14.2%±2.33 Nov 2025, Terminal-Bench v2 Leaderboard
- effort unknown; Agent: Mini-SWE-Agent14.2%±2.35 Aug 2025, tbench.ai
Epoch's name: gpt-oss-120b. Best closed model on this set: gpt-5.5_unknown, 84.7% (OpenAI). All 202 results
- effort high48.2%WeirdML Leaderboard
- effort high48.2%±0.0WeirdML Leaderboard, via Epoch AI
- effort medium41.9%±0.0htihle.github.io
Epoch's name: gpt-oss-120b, openai/gpt-oss-120b. Best closed model on this set: gpt-6-astra_promax, 93.6% (OpenAI). All 173 results
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Independent scores from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Each row links its own source. What the badges mean.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 0.29 GB, worked out from the model's config (openai/gpt-oss-120b), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
ggml-org/gpt-oss-120b-GGUF (converted by ggml-org), 391k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| MXFP4 | 59.0 GBfile size | 59.8 GB59.0 + 0.29 + 0.50 | Commandsllama.cpp llama-server -hf ggml-org/gpt-oss-120b-GGUF:MXFP4Ollama ollama run hf.co/ggml-org/gpt-oss-120b-GGUF:MXFP4Download hf download ggml-org/gpt-oss-120b-GGUF --include "gpt-oss-120b-MXFP4.gguf" --local-dir gpt-oss-120b-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX mxfp4mlx-community/gpt-oss-120b-MXFP4-Q8 | 59.0 GBfile size | 59.8 GB59.0 + 0.29 + 0.50 | Download page |
| MLX 8bitlmstudio-community/gpt-oss-120b-MLX-8bit | 115.6 GBfile size | 116.4 GB115.6 + 0.29 + 0.50 | Download page |
Original weights
The maker's own weights are packed into integer tensors, so the safetensors header gives no size we can trust; the repo lists the real file sizes.
hf download openai/gpt-oss-120b --local-dir gpt-oss-120bOther conversions (9; sizes not read)
- unsloth/gpt-oss-120b-unsloth-bnb-4bit bitsandbytes, 4bit, 4.5k downloads
- amd/gpt-oss-120b-w-mxfp4-a-fp8 FP8, mxfp4, 3.5k downloads
- furiosa-ai/gpt-oss-120b Other, 1.4k downloads
- unsloth/gpt-oss-120b Other, 1.3k downloads
- xCloudinfo/gpt-oss-120b-TAIDE-zhTW Other, 331 downloads
- RedHatAI/gpt-oss-120b Other, 234 downloads
- positron-ai/openai_gpt-oss-120b-ingest-best-gptq GPTQ, 89 downloads
- amd/gpt-oss-120b-w-mxfp4-a-fp8-kv-fp8-fp8attn-no_lmhead_router FP8, mxfp4, 44 downloads
- amd/gpt-oss-120b-w-mxfp4-a-fp8-Mlperf FP8, mxfp4, 40 downloads
The Hub tags 100 or more quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
22 providers serve this model through the Hugging Face router. Prices are US dollars per million tokens. Precision is what the provider reports; a lower precision than the original weights can change answers. Live first, then cheapest output.
| Provider | In | Out | Context | Precision | Tokens/s | First token | Uptime 24h | Tools | JSON |
|---|---|---|---|---|---|---|---|---|---|
| Near AI | $0.03 | $0.15 | 128k | fp4 | 84 | 0.5 s | 99.9% | yes | yes |
| Venice | $0.03 | $0.15 | 125k | int4 | 53 | 1.0 s | 100.0% | yes | no |
| deepinfra | $0.037 | $0.17 | 128k | not given | 26 | 0.3 s | ? | yes | yes |
| CoreWeave | $0.03 | $0.17 | 128k | fp4 | 139 | 0.4 s | 98.9% | yes | yes |
| AkashML | $0.03 | $0.19 | 128k | fp4 | 127 | 0.5 s | 100.0% | yes | yes |
| Crusoe | $0.05 | $0.25 | 128k | bf16 | 212 | 0.3 s | 100.0% | yes | yes |
| novita | $0.05 | $0.25 | 128k | not given | 177 | 0.6 s | ? | yes | yes |
| $0.09 | $0.36 | 128k | not given | 230 | 0.3 s | 99.7% | no | yes | |
| Mancer 2 | $0.095 | $0.38 | 128k | fp8 | 116 | 0.6 s | 99.0% | yes | yes |
| Sail Research | $0.06 | $0.40 | 128k | fp4 | 130 | 0.8 s | 100.0% | yes | yes |
| DigitalOcean | $0.06 | $0.42 | 125k | not given | 58 | 0.9 s | 100.0% | no | no |
| ovhcloud | $0.09 | $0.47 | 128k | not given | 34 | 0.4 s | ? | yes | yes |
| BaseTen | $0.10 | $0.50 | 125k | fp4 | 178 | 0.2 s | 99.7% | yes | yes |
| Amazon Bedrock | $0.15 | $0.60 | 128k | not given | 163 | 0.4 s | 99.8% | yes | no |
| Groq | $0.15 | $0.60 | 128k | not given | 343 | 0.2 s | 99.5% | yes | yes |
| Nebius | $0.15 | $0.60 | 128k | fp4 | 194 | 0.3 s | 100.0% | yes | yes |
| SiliconFlow | $0.15 | $0.60 | 128k | fp8 | 63 | 1.1 s | 100.0% | no | yes |
| Together | $0.15 | $0.60 | 128k | not given | 74 | 0.3 s | 99.9% | yes | yes |
| Cerebras | $0.35 | $0.75 | 128k | fp16 | 719 | 0.2 s | 100.0% | yes | yes |
| Parasail | $0.10 | $0.75 | 128k | fp4 | 186 | 0.5 s | 100.0% | yes | yes |
| SambaNova | $0.14 | $0.95 | 128k | not given | 326 | 0.4 s | 99.3% | yes | no |
| featherless-ai | ? | ? | ? | not given | ? | ? | ? | not given | not given |
Through the router, with a Hugging Face token (the router bills your Hugging Face account). Change the name after the colon to pick another provider.
curl https://router.huggingface.co/v1/chat/completions \
-H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" \
-d '{"model": "openai/gpt-oss-120b:near-ai", "messages": [{"role": "user", "content": "Hello"}]}'Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Fine-tunes of this model
Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.
- nvidia/gpt-oss-120b-Eagle3-short-context206kUnreviewed
- yujiepan/gpt-oss-tiny-random7M42kUnreviewed
- openai/gpt-oss-safeguard-120b20kUnreviewed
- nvidia/gpt-oss-120b-Eagle3-v33.5kUnreviewed
- nvidia/gpt-oss-120b-Eagle3-long-context1.5kUnreviewed
- nvidia/gpt-oss-120b-Eagle3-throughput678Unreviewed
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.