GPT-OSS 120B gpt-oss-120b

openai · gpt-oss reasoning flagship efficient

#35 overall #6 coding #21 agentic

params
117B (A5.1B)
arch
moe
context
128k
license
apache-2.0 open weights
released
Aug 2025
reasoning
yes
train compute
4.9e+24 FLOP
downloads/30d
5.3M

# architecture

attn shared · reasoning GQA 64q/8kv router top-4 of 128 expert ×128 expert out ×36 layers ctx 131,072 117B pool · A5.1B/token
layers
36
d_model
2880
heads
64q / 8kv
head dim
64
experts
top-4 of 128
vocab
201088
family
gpt_oss

signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.

# why ranked

Overall: #35 — score 45.7 ● high — all signals present

signalweightinput (0–100)
aa_intelligence 0.70 53.1
bench_composite 0.30 28.4

benchmark panel evidence:

complete panelscore (0–100)
aider-polyglot-v137.6
swe-bench-verified-v119.2

Coding: #6 — score 24.2 ● high — all signals present

signalweightinput (0–100)
aa_coding 0.30 35.1
aider_polyglot 0.30 45.5
swe_bench_verified 0.40 0.0

Agentic: #21 — score — ● low — a single signal; treat with caution

signalweightinput (0–100)
swe_bench_verified 0.60 0.0
swe_rebench 0.40

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

OGM score · last 55 days
OGM score over 55 days
overall rank (up is better)
Overall rank over 55 days

methodology changed during this history window; score movement across that boundary is not model movement. See methodology v5.

# benchmarks

benchmarkscoresourcedate
AA Coding Index 30.4 Artificial Analysis
AA Intelligence Index 15.6 Artificial Analysis
AA Math Index 93.4 Artificial Analysis
Aider Polyglot 41.8 Aider Aug 2025
Codeforces 0.8 / 3000 LLM Stats
SWE-bench Verified 26.0 SWE-bench Aug 2025

# where to run

providerquantctx$/M in$/M out$/M cacheprice srctpsuptime
Featherless via hfrouter
AkashMLbf16 128k $0.03$0.17$0.03 via openrouter 99.8%
CoreWeavefp4 128k $0.03$0.17$0.03 via openrouter 99.9%
Wandb 128k $0.03$0.17 via litellm
DeepInfrabf16 128k $0.04$0.17 deepinfra 99.9%
DeepInfrabf16 128k $0.04$0.17 deepinfra 100.0%
DeepInfrafp8 128k $0.04$0.17 deepinfra 98.7%
Novitafp4 128k $0.05$0.25 novita 96.2%
Venicenot-available 125k $0.07$0.30 venice
Amazon Bedrockunknown 128k $0.07$0.30 bedrock 100.0%
Amazon Bedrockunknown 128k $0.07$0.30 bedrock 93.1%
DigitalOceanunknown 125k $0.06$0.39$0.02 via openrouter 100.0%
SiliconFlowfp8 128k $0.05$0.45 openrouter 90.2%
Googleunknown 128k $0.09$0.36 via openrouter 90.5%
Mancer 2fp8 128k $0.06$0.50 via openrouter 99.4%
Nscale 128k $0.10$0.40 via hfrouter 113
OVHcloud 128k $0.09$0.47 ovhcloud 68
BaseTenfp4 128k $0.10$0.50$0.10 via openrouter 100.0%
BaseTenfp4 128k $0.10$0.50$0.10 via openrouter 100.0%
Azure $0.15$0.60 azure
Bedrock Mantle 128k $0.15$0.60 via litellm
Fireworks 128k $0.15$0.60 via hfrouter 180
Groqunknown 128k $0.15$0.60 groq 100.0%
Nebiusfp4 128k $0.15$0.60 openrouter 97.6%
Phalaunknown 128k $0.15$0.60 phala 97.5%
Tensormesh 128k $0.15$0.60 via litellm
Togetherunknown 128k $0.15$0.60 via openrouter 95.3%
Parasailfp4 128k $0.10$0.75$0.06 via openrouter 100.0%
IBM watsonx 128k $0.16$0.64 via modelsdev
Scaleway $0.17$0.68 via hfrouter 75
Maraunknown 128k $0.15$0.75 via openrouter 93.0%
SambaNovaunknown 128k $0.22$0.59 sambanova 99.5%
Replicate $0.18$0.72 via litellm
Cerebrasfp16 128k $0.35$0.75$0.35 openrouter 100.0%
Cloudflare 125k $0.35$0.75 via modelsdev
Crusoe 128k $0.80$0.80 via litellm

sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price

source aliases
aa
gpt-oss-120b
aider
gpt-oss-120b (high)
arena
gpt-oss-120b
openrouter
openai/gpt-oss-120b