OpenAI: gpt-oss-120b

by OpenAI

gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized to run on a single H100 GPU with native MXFP4 quantization. The model supports configurable reasoning depth, full chain-of-thought access, and native tool use, including function calling, browsing, and structured output generation.

Avg Score

48.3%

259 answers

Avg Latency

11.2s

25 runs

Pricing

Free

input

Free

output

per 1M tokens

Context

131K

tokens

Alternatives

Models with similar or better quality but different tradeoffs

Same Quality, Cheaper

Models with similar or better performance at a lower cost per token.

Model	Cost
Meta: Llama 3.2 3B Instruct	-98%
Meta: Llama 4 Scout	-79%
Meta: Llama 3.3 70B Instruct	-78%
Mistral: Ministral 3 3B 2512	-70%
Google: Gemini 2.0 Flash	-68%

Same Quality, Faster

Models with similar or better performance but lower latency.

Model	Latency
Google: Gemini 2.5 Flash Lite	-90%
Google: Gemini 2.0 Flash	-89%
Google: Gemini 2.5 Flash	-87%
Mistral: Ministral 3 8B 2512	-85%
Mistral: Ministral 3 3B 2512	-85%

Same Cost, Better

Models at a similar price point with higher benchmark scores.

Model	Score
DeepSeek: DeepSeek V3 0324	+31%
DeepSeek: DeepSeek V3.2	+31%
Google: Gemini 2.0 Flash	+30%
Anthropic: Claude 3 Haiku	+27%
Meta: Llama 4 Maverick	+27%

Other Models from OpenAI

Compare performance with other models from the same creator

Model	Score	Latency	Cost/1M
OpenAI: GPT-5.2	87.1%	4.9s	$7.88
OpenAI: GPT-5.2 Pro	85.5%	18.4s	$94.50
OpenAI: GPT-5.2 Chat	84.9%	4.2s	$7.88
OpenAI: GPT-5 Mini	80.1%	8.9s	$1.13
OpenAI: GPT-5 Nano	78.7%	13.3s	$0.22
OpenAI: gpt-oss-20b	73.3%	6.4s	$0.06
OpenAI: GPT-5.1-Codex-Max	—	—	$5.63
OpenAI: GPT-5.1	—	—	$5.63
OpenAI: GPT-5.1 Chat	—	—	$5.63
OpenAI: GPT-5.1-Codex	—	—	$5.63
OpenAI: GPT-5.1-Codex-Mini	—	—	$1.13
OpenAI: gpt-oss-safeguard-20b	—	—	$0.19
OpenAI: GPT-5 Image Mini	—	—	$2.25
OpenAI: GPT-5 Image	—	—	$10.00
OpenAI: o3 Deep Research	—	—	$25.00
OpenAI: o4 Mini Deep Research	—	—	$5.00
OpenAI: GPT-5 Pro	—	—	$67.50
OpenAI: GPT-5 Codex	—	—	$5.63
OpenAI: GPT-4o Audio	—	—	$6.25
OpenAI: GPT-5 Chat	—	—	$5.63
OpenAI: GPT-5	—	—	$5.63
OpenAI: gpt-oss-20b	—	—	$0.0000
OpenAI: o3 Pro	—	—	$50.00
OpenAI: Codex Mini	—	—	$3.75
OpenAI: o4 Mini High	—	—	$2.75
OpenAI: o3	—	—	$5.00
OpenAI: o4 Mini	—	—	$2.75
OpenAI: GPT-4.1	—	—	$5.00
OpenAI: GPT-4.1 Mini	—	—	$1.00
OpenAI: GPT-4.1 Nano	—	—	$0.25
OpenAI: o1-pro	—	—	$375.00
OpenAI: GPT-4o-mini Search Preview	—	—	$0.38
OpenAI: GPT-4o Search Preview	—	—	$6.25
OpenAI: o3 Mini High	—	—	$2.75
OpenAI: o3 Mini	—	—	$2.75
OpenAI: o1	—	—	$37.50
OpenAI: GPT-4o (2024-11-20)	—	—	$6.25
OpenAI: ChatGPT-4o	—	—	$10.00
OpenAI: GPT-4o (2024-08-06)	—	—	$6.25
OpenAI: GPT-4o-mini (2024-07-18)	—	—	$0.38
OpenAI: GPT-4o-mini	—	—	$0.38
OpenAI: GPT-4o (2024-05-13)	—	—	$10.00
OpenAI: GPT-4o	—	—	$6.25
OpenAI: GPT-4o	—	—	$12.00
OpenAI: GPT-4 Turbo	—	—	$20.00
OpenAI: GPT-3.5 Turbo (older v0613)	—	—	$1.50
OpenAI: GPT-4 Turbo Preview	—	—	$20.00
OpenAI: GPT-4 Turbo (older v1106)	—	—	$20.00
OpenAI: GPT-3.5 Turbo Instruct	—	—	$1.75
OpenAI: GPT-3.5 Turbo 16k	—	—	$3.50
OpenAI: GPT-4 (older v0314)	—	—	$45.00
OpenAI: GPT-4	—	—	$45.00
OpenAI: GPT-3.5 Turbo	—	—	$1.00

Benchmark Performance

How this model performs across different benchmarks

Benchmark	Score	Rank
KUKA force Torque control programming skills	100.0%	2 / 17
Customer service	100.0%	15 / 17
Customer support	100.0%	6 / 17
Structured Customer Support Extraction	100.0%	2 / 17
solve n-queens in clojure in one shot	100.0%	6 / 16
Repetitive Onomatopoeia Response Benchmark	100.0%	15 / 17
Motor production test	100.0%	12 / 36
SaaS Customer Support Extraction and Strategy	100.0%	4 / 17
Venture Capital Terms Benchmark	94.4%	16 / 25
Spatial Reasoning: Germany	93.2%	15 / 35
Character Frequency Bench	90.9%	15 / 35
Product Recommendations	87.5%	3 / 36
Karlsruhe Local Knowledge Benchmark	60.8%	28 / 35
Math Bench: Arithmetic (1-1000)	45.0%	33 / 35
Math Bench: Arithmetic (1-100)	42.0%	33 / 35
Math Bench: Division (1-10)	37.0%	34 / 35
Money Boy Cultural Literacy Test	32.9%	16 / 35
Math Bench: Addition (1-10)	32.0%	34 / 35
Math Bench: Subtraction (1-10)	27.0%	34 / 35
Math Bench: Multiplication (1-10)	26.0%	34 / 35
German Memelord Bench	9.0%	25 / 35
Karl Lorey knowlege	0.0%	32 / 36
Current Global Political Leaders	0.0%	11 / 17

Price vs Performance

Compare cost efficiency across all models

Current model

Other models

X-axis uses log scale for better visualization of price range

Score Over Time

Performance trends across all benchmark runs

Benchmark Activity

Number of benchmark runs over time

Quickstart

Get started with this model using OpenRouter

View on OpenRouter

import { OpenRouter } from "@openrouter/sdk";

const openrouter = new OpenRouter({
  apiKey: "<OPENROUTER_API_KEY>"
});

const completion = await openrouter.chat.completions.create({
  model: "openai/gpt-oss-120b:free",
  messages: [
    {
      role: "user",
      content: "Hello!"
    }
  ]
});

console.log(completion.choices[0].message.content);

Get your API key at openrouter.ai/keys