← Model list

OpenAI GPT OSS 120B

openai·openai/gpt-oss-120b-11·GA·Open weights·gpt-oss series·NEW

Open-weight GPT model for self-hosted reasoning and instruction-following workloads

At a glance
Reference price$2.92 / $2.92 per 1M
Cache read · CloudFerro Sherlock
Lowest paid$0.04 / $0.16
Helicone Gateway · 73× spread
Context131,000
⚠ Providers report 128,000–131,072; the table below is authoritative
Output limit131,000
Capabilities
ReasoningTool useStructured outputTemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
TextImage
Knowledge cutoff2024-06
Released / updated2025-08-28 / 2026-06-11

Quality & performance

Artificial Analysis doesn't cover this model (267 of 2059 have data). Quality data comes from independent evals covering widely used models.

Available at 6 providers6 with public prices

ProviderTierInputOutputCache readCache writeContextOutput limitStatus
Helicone
gpt-oss-120b
Gateway$0.04$0.16131,072131,072
NovitaAI
openai/gpt-oss-120b
Gateway$0.05$0.25131,07232,768
Venice AI
openai-gpt-oss-120b
Gateway$0.07$0.30128,00016,384
DigitalOcean
openai-gpt-oss-120b
Cloud$0.055$0.385$0.02128,0004,096
SiliconFlow
openai/gpt-oss-120b
Cloud$0.05$0.45131,0008,000
CloudFerro Sherlock
openai/gpt-oss-120b
Gateway$2.92$2.92131,000131,000

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Your usage cost

1Helicone$16.00
2NovitaAI$22.50
3DigitalOcean$26.05
4Venice AI$29.00
5SiliconFlow$32.50
6CloudFerro Sherlock$730.00
The cheapest paid channel is the only channel.

Benchmark

No upstream benchmark data for this model. For quality, see the Artificial Analysis intelligence score above.

Reasoning control

effort = loweffort = mediumeffort = higheffort = nonebudget_tokens ≥ 128 ≤ 32,768

Related models

Price historyone sample accumulated per data sync

Input list $2.92Output list $2.92Min blended $0.07

Price history accumulates from each data sync; currently only 1 sample(s) (2026-08-21). Each future sync adds a point, and once accumulated a line is drawn here.