gpt-oss-120b
Where it stands
Quality in independent tests against price, next to 135 other models. Up means more capable, left means cheaper.
About this model
gpt-oss-120b is an AI model from OpenAI, made available through the API on August 5, 2025. It can read up to 131,072 tokens in one request, roughly 246 pages of text, and write up to 117,964 tokens in a single answer.
Through the API it costs $0.037 per million input tokens and $0.17 per million output tokens. Its pricing puts it among the more affordable models we track, which makes it a good fit for high-volume tasks.
It supports extended reasoning: it can "think" step by step before answering, which helps on maths, code and complex questions. It can call tools and functions, so developers can connect it to search, databases or other software. Its weights are openly published (openai/gpt-oss-120b), so it can also be downloaded and run on your own hardware.
Quality scores
Rankings →- GPQA DiamondVery hard multiple-choice questions in biology, physics and chemistry, written by PhD experts. Random guessing scores 25%.75.8%#75 of 116
- AIME mathCompetition-level math problems in the style of the American Invitational Mathematics Examination. Share solved.88.9%#45 of 112
Where to use it
Companies that serve this model through an API: 24. Prices per million tokens, cheapest first.
| Provider | Inputi | Outputi | Contexti | Uptime (24 h)i |
|---|---|---|---|---|
CoreWeaveCheapestfp4 | $0.03 | $0.17 | 128K | 99.6% |
DeepInfrabf16 | $0.037 | $0.17 | 128K | 98.9% |
DekaLLMbf16 | $0.03 | $0.18 | 128K | 99.1% |
AkashMLbf16 | $0.037 | $0.19 | 128K | 100.0% |
Crusoebf16 | $0.05 | $0.25 | 128K | 100.0% |
Novitafp4 | $0.05 | $0.25 | 128K | 99.5% |
Mancer 2fp8 | $0.05 | $0.30 | 128K | 98.8% |
Google | $0.09 | $0.36 | 128K | 52.3% |
DigitalOcean | $0.06 | $0.42 | 128K | 99.9% |
BaseTenfp4 | $0.10 | $0.50 | 128K | 99.4% |
BaseTenfp4 | $0.10 | $0.50 | 128K | 99.6% |
Amazon Bedrock | $0.15 | $0.60 | 128K | 98.7% |
Nebiusfp4 | $0.15 | $0.60 | 128K | 92.1% |
Amazon Bedrock | $0.15 | $0.60 | 128K | 100.0% |
DeepInfrabf16 | $0.15 | $0.60 | 128K | 100.0% |
SiliconFlowfp8 | $0.15 | $0.60 | 128K | 82.8% |
Phala | $0.15 | $0.60 | 128K | 99.6% |
Together | $0.15 | $0.60 | 128K | 97.6% |
Groq | $0.15 | $0.60 | 128K | 99.3% |
Parasailfp4 | $0.10 | $0.75 | 128K | 100.0% |
Mara | $0.15 | $0.75 | 128K | 64.4% |
SambaNova | $0.14 | $0.95 | 128K | 97.6% |
Cerebrasfp16 | $0.35 | $0.75 | 128K | 100.0% |
DeepInfrafp8 | $0.20 | $0.95 | 128K | 91.7% |
Availability and prices come from OpenRouter’s public API and are refreshed daily. Uptime is the share of successful requests over the last 24 hours. Some providers run a compressed version of open models (fp8, int4), which is cheaper but can be slightly less accurate.
Use it in your code
Call gpt-oss-120b through OpenRouter’s OpenAI-compatible API: create a key on openrouter.ai, then run one of these examples.
curl https://openrouter.ai/api/v1/chat/completions \
-H "Authorization: Bearer $OPENROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "openai/gpt-oss-120b", "messages": [{"role": "user", "content": "Hello! What can you do?"}]}'import os
from openai import OpenAI
client = OpenAI(base_url="https://openrouter.ai/api/v1", api_key=os.environ["OPENROUTER_API_KEY"])
reply = client.chat.completions.create(
model="openai/gpt-oss-120b",
messages=[{"role": "user", "content": "Hello! What can you do?"}],
)
print(reply.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({ baseURL: "https://openrouter.ai/api/v1", apiKey: process.env.OPENROUTER_API_KEY });
const reply = await client.chat.completions.create({
model: "openai/gpt-oss-120b",
messages: [{ role: "user", content: "Hello! What can you do?" }],
});
console.log(reply.choices[0].message.content);The same code works with every model on this site: only the model ID changes. Many labs also offer their own API and SDK.
What it costs in practice
- 1,000 chat replies$0.07
- Summarise a 100-page document$0.002
- An app with 10,000 requests a day for a month$31.50
Similar-priced alternatives
| Model | Inputi | Outputi | Contexti | |
|---|---|---|---|---|
Mercury 2.5Inception | $0.04 | $0.15 | 260K | Compare with ⇄ |
Hy-MT2-1.8BTencent | $0.044 | $0.18 | 8K | Compare with ⇄ |
Schematron V2 TurboInference Net | $0.03 | $0.15 | 128K | Compare with ⇄ |
Qwen3.7 FlashAlibaba Qwen | $0.03 | $0.13 | 1M | Compare with ⇄ |
Solar Mini 4Upstage | $0.05 | $0.20 | 512K | Compare with ⇄ |
Nemotron 3 Nano 30B A3BNVIDIA | $0.05 | $0.20 | 256K | Compare with ⇄ |
News about OpenAI
See all →Frequently asked questions
- How much does gpt-oss-120b cost?
- gpt-oss-120b costs $0.037 per million input tokens and $0.17 per million output tokens through the API. For example, 1,000 chat replies of about 500 tokens in and 300 out cost around $0.07.
- What is the context window of gpt-oss-120b?
- gpt-oss-120b can read up to 131,072 tokens in one request, about 246 pages of text. It can write up to 117,964 tokens in a single answer.
- Is gpt-oss-120b open source?
- Yes. Its weights are published (openai/gpt-oss-120b), so you can download it and run it on your own hardware, or use it through one of 24 API providers.
- Who makes gpt-oss-120b?
- gpt-oss-120b is made by OpenAI. It became available through the API on August 5, 2025.
- How good is gpt-oss-120b?
- gpt-oss-120b has an overall index of 139.9 in Epoch AI’s independent tests, ranking #96 of 136 models we track.
- Can gpt-oss-120b read images?
- No. gpt-oss-120b only accepts text as input.
- What are the alternatives to gpt-oss-120b?
- Models with a similar price from other companies include Mercury 2.5, Hy-MT2-1.8B and Schematron V2 Turbo. You can compare them side by side on this page.