OpenAI › GPT OSS 120B › Pricing
$0.03 / $0.15per 1M tokens in / out
#43on OpenRouter by tokens · 514.4B / week
Price sheetper 1M tokens
| Input | $0.03 | Near AI |
| Output | $0.15 | |
| Context | 131K | tokens |
Where to buy24 providers · per 1M tokens
| Provider | Input | Output | Context |
|---|---|---|---|
| AkashML | $0.03 | $0.19 | — |
| CoreWeave | $0.03 | $0.17 | — |
| DekaLLM | $0.03 | $0.18 | — |
| Near AI | $0.03 | $0.15 | — |
| Venice | $0.03 | $0.15 | — |
| DeepInfra | $0.037 | $0.17 | — |
| Crusoe | $0.05 | $0.25 | — |
| Mancer 2 | $0.05 | $0.25 | — |
| Novita | $0.05 | $0.25 | — |
| DigitalOcean | $0.06 | $0.42 | — |
| Azure AI Foundry | $0.082 | $0.66 | — |
| $0.09 | $0.36 | 131K | |
| BaseTen | $0.10 | $0.50 | — |
| Parasail | $0.10 | $0.75 | — |
| SambaNova | $0.14 | $0.95 | — |
| Amazon Bedrock | $0.15 | $0.60 | — |
| AWS Bedrock | $0.15 | $0.60 | — |
| Featherless | $0.15 | $0.60 | — |
| Groq | $0.15 | $0.60 | — |
| Nebius | $0.15 | $0.60 | — |
| Phala | $0.15 | $0.60 | 131K |
| SiliconFlow | $0.15 | $0.60 | — |
| together | $0.15 | $0.60 | — |
| Cerebras | $0.35 | $0.75 | — |
In OpenAI's line-upper 1M in / out
| GPT-6 Luna Decisions | $0.10 / — | −100% |
| GPT OSS 20B | $0.018 / $0.09 | −40% |
| GPT OSS 120B | $0.03 / $0.15 | this model |
| GPT OSS Safeguard 20B | $0.07 / $0.20 | +71% |
| GPT-6 Luna | $0.10 / $0.50 | 3.3× |
| GPT OSS Safeguard 120B | $0.15 / $0.60 | 4.4× |
| GPT-5.4 Nano (2026-03-17) | $0.20 / $1.25 | 7.7× |
Cheaper for the same workvs GPT OSS 120B at low
| Nemotron 3.5 Lightning thinkingOpen | -19% | score +26% | $0.079 |
| GPT OSS 20B lowOpen | -40% | same score | $0.018 / $0.09 |
| Granite 4.2 3B thinkingOpen | -13% | score -11% | $0.006 |
| Gemma 4 E4B thinkingOpen | -33% | score -13% | $0.008 |
| Qwen3.5 2B thinkingOpen | -33% | score -32% | $0.02 / $0.10 |
Pricing questions
How much does GPT OSS 120B cost?›
$0.03 per 1M input tokens and $0.15 per 1M output tokens.
Is GPT OSS 120B free to use?›
There is no free API tier, but the weights are open: you can run it on your own hardware for the cost of compute.
What is a cheaper alternative to GPT OSS 120B?›
Nemotron 3.5 Lightning at thinking effort costs -19% per task with a score +26% on Humanity's Last Exam.