Model FatigueLLMs under $1, measured — and who holds your data
Models

gpt-oss-20b

GPT OSS 20B

Made by OpenAI. Sold by 3 providers, from ≈$0.04 per million input tokens.

Maker
OpenAI
Released
2025-08-05
Context
131K tokens
Max output
131K tokens
Input
text
Output
text
Open weights
Yes
Knowledge cutoff
2024-10

Where to buy it

ProviderInput $/MOutput $/MBlended*Holds your promptTraining
OVHcloud≈$0.04
€0.04
≈$0.17
€0.15
≈$0.08runs it 🇫🇷 Franceno training
Amazon BedrockOpenAI-compatible API$0.07$0.30$0.13runs it 🇺🇸 United Statesno training
Amazon Bedrockruntime API$0.07$0.30$0.13runs it 🇺🇸 United Statesno training
Google Vertex AI$0.07$0.25$0.12runs it 🇺🇸 United Statesoff by default

≈ converted from EUR at 1 EUR = $1.1193 (ECB reference rate, 2026-10-23)

Blended is our calculation: (3 × input + output) ÷ 4, US dollars per million tokens — most calls read far more than they write. Tiers that are cheaper because they train on your data are not used for the “from” price.

Compare with every other model →