gpt-oss 120B on a 192GB AMD mini PC "at 55.7 tps"
CodeRead the post
x.com · link re-checked 2026-10-16
GPT Oss 120B
By OpenAI. 4 results on 4 boards, copied as published. Different boards, different metrics: don't add them up.
License (as listed)Not listed
Hosting regionNot confirmed
Boards4
Code
- 33.6%Vals AI — SWE-bench Verified
GPT OSS 120B · % resolved · board #87
Mini-SWE-agent · as of 2026-09-01 Stale ↗
- 83.23%Vals AI — LiveCodeBench
GPT OSS 120B · Accuracy % · board #51
Vals AI implementation (no agent) · as of 2026-09-01 Stale ↗
- 16.2%SWE-Bench Pro (Scale, public set)
gpt-oss-120b · % resolved · board #19
SWE-Agent · as of 2026-10-16 ↗
Agentic / terminal
No result on the boards we captured
Websites
No result on the boards we captured
Design
- 961Design Arena (All categories)
gpt-oss-120b · Elo · board #173
Design Arena · as of 2026-10-24 ↗
Android
No result on the boards we captured
iOS
No mature public leaderboard
Available in
No plan row in our pricing data names this model version.
Use cases with this model
"I JUST RAN GPT OSS 120B FOR $0" via OpenCode → local endpoint
CodeRead the post
x.com · link re-checked 2026-10-16
Merrymen agent moving off gpt-oss-120b to its own model
CodeRead the post
x.com · link re-checked 2026-10-16
Same family: gpt-oss 1
Unreal Tournament 99 bot driven by "gpt-oss-20b running on my own GPUs"
CodeRead the post
x.com · link re-checked 2026-10-16
Sources
vals.ai/benchmarks/swebench vals.ai/benchmarks/lcb scale.com/leaderboard/swe_bench_pro_public designarena.ai/leaderboard