GPT OSS 120B API Pricing
OpenAI · released 2025-08-05
GPT OSS 120B price
No official API price published for GPT OSS 120B yet.
Benchmarks
| Benchmark | Value | Source | As of |
|---|---|---|---|
| Epoch Capabilities Index (ECI) | 140.2 | epoch | 2025-08-05 |
| Epoch Mock AIME accuracy | 88.9% | epoch | 2025-08-05 |
| Epoch FrontierMath (T1–3) accuracy | No public data yet | ||
| Arena text rating | 1352 | arena | 2026-10-28 |
| Arena coding rating (human preference votes, not task accuracy) | 1389 | arena | 2026-10-28 |
| Arena math rating | 1381 | arena | 2026-10-28 |
| Arena WebDev rating | No public data yet | ||
More from OpenAI
| Model | Released | Blended $/1M |
|---|---|---|
| GPT-6.1 Terra | 2026-10-26 | $2.00 |
| GPT-6.1 Sol | 2026-09-29 | $4.00 |
| GPT-6 Sol | 2026-09-22 | $4.00 |
| GPT-6 Luna | 2026-09-22 | $0.20 |
| GPT-Live 1 | 2026-09-10 | — |
| GPT Image 2.5 Sunburst | 2026-09-08 | $11.25 |
| GPT Image 2.5 Flare | 2026-09-08 | $11.25 |
| GPT-6 Astra (Fast) | 2026-09-04 | — |
See GPT OSS 120B on the leaderboard
Prices: OpenAI's official list price as recorded by None; benchmarks: see sources below; last synced 2026-11-06.
FAQ
Who develops GPT OSS 120B?
GPT OSS 120B is developed by OpenAI, released 2025-08-05.
Is GPT OSS 120B's API pricing published?
No — as of the last data sync (2026-11-06), neither models.dev nor LiteLLM has published official API pricing for GPT OSS 120B. This page updates automatically once pricing is available.
How does GPT OSS 120B rank on benchmarks?
GPT OSS 120B scores 140.2 on the Epoch Capabilities Index (ECI), ranking #88 of 107 models we track with a published ECI score.
Which benchmarks is GPT OSS 120B scored on?
GPT OSS 120B has published results for: Epoch Capabilities Index (ECI); Epoch Mock AIME accuracy; Arena text rating; Arena coding rating (human preference votes, not task accuracy); Arena math rating. The Benchmarks table on this page shows each score with its source and date.
Where does this data come from?
Pricing for GPT OSS 120B comes from models.dev and LiteLLM (open MIT-licensed catalogs); benchmark scores come from Epoch AI and LMArena (open CC BY 4.0 licenses) — see /data-sources/. Last synced 2026-11-06.