Skip to content
Weights Not Walls
Menu

openai

gpt-oss-20b

Reviewed
Size
20.9B parametersfrom the GGUF metadata of a conversion, because this repo's own weights are packed: unsloth/gpt-oss-20b-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-20b5.6M downloads in the last 30 days, 5.3k likes

Scores

16 independent results on 7 eval sets
GPQA diamondRe-run by Epoch AI
  • effort medium60.8%±2.527 Aug 2026, log
  • effort low53.2%±2.527 Aug 2026, log
  • effort high46.0%±3.66 Aug 2026, log

Epoch's name: gpt-oss-20b. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results

OTIS Mock AIME 2024-2025Re-run by Epoch AI
  • effort medium65.3%±5.927 Aug 2026, log
  • effort high50.8%±6.727 Aug 2026, log
  • effort low40.3%±6.327 Aug 2026, log

Epoch's name: gpt-oss-20b. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 results

CritPtReported

Epoch's name: gpt-oss-20b. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results

DTBenchReported

Epoch's name: gpt-oss-20b. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results

LMCAReported

Epoch's name: gpt-oss-20b. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results

Epoch's name: gpt-oss-20b. Best closed model on this set: gpt-5.5_unknown, 84.7% (OpenAI). All 202 results

WeirdMLReported

Epoch's name: gpt-oss-20b. 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

file sizes read from the Hub 19 Nov 2026

Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 0.19 GB, worked out from the model's config (openai/gpt-oss-20b), 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

unsloth/gpt-oss-20b-GGUF (converted by unsloth), 608k downloads. Sizes are the real files on the Hub, all parts added up.

FormatSizeMemory neededGet it
Q3_K_S10.7 GBfile size11.4 GB10.7 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q3_K_S
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q3_K_S
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q3_K_S.gguf" --local-dir gpt-oss-20b-GGUF
Q4_010.7 GBfile size11.4 GB10.7 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q4_0
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q4_0
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q4_0.gguf" --local-dir gpt-oss-20b-GGUF
Q3_K_M10.7 GBfile size11.4 GB10.7 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q3_K_M
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q3_K_M
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q3_K_M.gguf" --local-dir gpt-oss-20b-GGUF
Q4_110.8 GBfile size11.5 GB10.8 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q4_1
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q4_1
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q4_1.gguf" --local-dir gpt-oss-20b-GGUF
Q4_K_S10.8 GBfile size11.5 GB10.8 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q4_K_S
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q4_K_S
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q4_K_S.gguf" --local-dir gpt-oss-20b-GGUF
Q4_K_M10.8 GBfile size11.5 GB10.8 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q4_K_M
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q4_K_M
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q4_K_M.gguf" --local-dir gpt-oss-20b-GGUF
Q5_K_S10.9 GBfile size11.6 GB10.9 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q5_K_S
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q5_K_S
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q5_K_S.gguf" --local-dir gpt-oss-20b-GGUF
Q5_K_M10.9 GBfile size11.6 GB10.9 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q5_K_M
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q5_K_M
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q5_K_M.gguf" --local-dir gpt-oss-20b-GGUF
UD-Q4_K_XL11.1 GBfile size11.7 GB11.1 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:UD-Q4_K_XL
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:UD-Q4_K_XL
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-UD-Q4_K_XL.gguf" --local-dir gpt-oss-20b-GGUF
Q6_K11.2 GBfile size11.9 GB11.2 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q6_K
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q6_K
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q6_K.gguf" --local-dir gpt-oss-20b-GGUF
UD-Q6_K_XL11.2 GBfile size11.9 GB11.2 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:UD-Q6_K_XL
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:UD-Q6_K_XL
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-UD-Q6_K_XL.gguf" --local-dir gpt-oss-20b-GGUF
Q8_011.3 GBfile size12.0 GB11.3 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:Q8_0
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:Q8_0
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-Q8_0.gguf" --local-dir gpt-oss-20b-GGUF
UD-Q8_K_XL12.3 GBfile size13.0 GB12.3 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:UD-Q8_K_XL
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:UD-Q8_K_XL
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-UD-Q8_K_XL.gguf" --local-dir gpt-oss-20b-GGUF
F1612.8 GBfile size13.5 GB12.8 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/gpt-oss-20b-GGUF:F16
Ollama
ollama run hf.co/unsloth/gpt-oss-20b-GGUF:F16
Download
hf download unsloth/gpt-oss-20b-GGUF --include "gpt-oss-20b-F16.gguf" --local-dir gpt-oss-20b-GGUF

MLX for Apple silicon

FormatSizeMemory neededRepo
MLX mxfp4mlx-community/gpt-oss-20b-MXFP4-Q811.2 GBfile size11.9 GB11.2 + 0.19 + 0.50Download page
MLX 8bitlmstudio-community/gpt-oss-20b-MLX-8bit20.7 GBfile size21.4 GB20.7 + 0.19 + 0.50Download 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-20b --local-dir gpt-oss-20b
Other conversions (7; sizes not read)

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

Hugging Face Inference Providers, fetched 19 Nov 2026, 13:58 UTC

10 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.

ProviderInOutContextPrecisionTokens/sFirst tokenUptime 24hToolsJSON
AkashML$0.018$0.09128kfp4720.6 s100.0%yesyes
CoreWeave$0.03$0.13128kfp4190.5 s99.7%yesyes
DeepInfra$0.03$0.14128kbf16720.2 s99.9%yesyes
Amazon Bedrock$0.07$0.15128knot given541.0 s99.9%yesno
Parasail$0.03$0.15128kfp4651.4 s99.8%yesyes
ovhcloud$0.05$0.18128knot given1380.4 s?yesyes
SiliconFlow$0.04$0.18128kfp8591.0 s87.6%noyes
Groq$0.075$0.30128knot given4530.4 s100.0%yesyes
PrimeIntellect$0.20$0.30128kfp42060.3 s99.9%yesyes
featherless-ai???not given???not givennot 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-20b:akashml", "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

from our snapshot of the Hub

Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.

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.