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

69.8%

741 answers

Avg Latency

29.0s

290 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 4 Scout	-64%
Mistral: Ministral 3 8B 2512	-59%
OpenAI: gpt-oss-20b	-46%
Mistral: Ministral 3 14B 2512	-36%
Google: Gemini 2.0 Flash	-36%

Same Quality, Faster

Models with similar or better performance but lower latency.

Model	Latency
Anthropic: Claude 3 Haiku	-81%
Google: Gemini 3 Flash Preview	-76%
Meta: Llama 4 Scout	-73%
Google: Gemini 2.0 Flash	-72%
Mistral: Ministral 3 8B 2512	-69%

Same Cost, Better

Models at a similar price point with higher benchmark scores.

Model	Score
DeepSeek: DeepSeek V3.2	+19%
DeepSeek: DeepSeek V3 0324	+18%
Meta: Llama 4 Maverick	+9%
Google: Gemini 2.0 Flash	+8%
Google: Gemini 2.5 Flash Lite	+7%

Other Models from OpenAI

Compare performance with other models from the same creator

Model	Score	Latency	Cost/1M
OpenAI: o1-pro	94.2%	95.3s	$375.00
OpenAI: GPT-5.1-Codex-Max	93.5%	25.2s	$5.63
OpenAI: GPT-5	92.2%	54.7s	$5.63
OpenAI: GPT-5 Image	90.9%	49.2s	$10.00
OpenAI: GPT-5.1 Chat	90.6%	6.1s	$5.63
OpenAI: GPT-5.1-Codex	90.6%	16.6s	$5.63
OpenAI: GPT-5.1-Codex-Mini	90.6%	12.6s	$1.13
OpenAI: GPT-5.1	90.3%	35.5s	$5.63
OpenAI: o3	90.3%	19.0s	$5.00
OpenAI: o4 Mini Deep Research	89.7%	134.5s	$5.00
OpenAI: GPT-5 Pro	89.4%	332.1s	$67.50
OpenAI: GPT-5 Image Mini	89.3%	31.8s	$2.25
OpenAI: GPT-4o Search Preview	89.3%	11.2s	$6.25
OpenAI: GPT-4.1	89.2%	21.8s	$5.00
OpenAI: gpt-oss-safeguard-20b	88.4%	2.3s	$0.19
OpenAI: o1	88.3%	24.1s	$37.50
OpenAI: GPT-5.2	87.8%	20.7s	$7.88
OpenAI: o3 Pro	87.6%	125.7s	$50.00
OpenAI: GPT-5.2-Codex	87.2%	21.7s	$7.88
OpenAI: GPT-5 Codex	87.2%	20.5s	$5.63
OpenAI: GPT-5.2 Pro	86.9%	47.7s	$94.50
OpenAI: o3 Deep Research	86.5%	362.7s	$25.00
OpenAI: GPT-5.2 Chat	86.4%	9.5s	$7.88
OpenAI: GPT-5 Chat	86.2%	6.8s	$5.63
OpenAI: o4 Mini High	86.1%	26.8s	$2.75
OpenAI: o3 Mini High	86.1%	13.8s	$2.75
OpenAI: o4 Mini	85.3%	18.2s	$2.75
OpenAI: o3 Mini	85.3%	16.0s	$2.75
OpenAI: ChatGPT-4o	82.9%	7.4s	$10.00
OpenAI: GPT-5 Mini	82.5%	24.7s	$1.13
OpenAI: GPT-5 Nano	77.1%	34.2s	$0.22
OpenAI: GPT-4o (2024-05-13)	76.5%	4.7s	$10.00
OpenAI: GPT-4o	76.1%	5.8s	$12.00
OpenAI: GPT-4o (2024-11-20)	75.6%	13.0s	$6.25
OpenAI: GPT-4o	73.8%	12.5s	$6.25
OpenAI: gpt-oss-20b	72.8%	13.3s	$0.06
OpenAI: GPT-4.1 Mini	71.9%	13.2s	$1.00
OpenAI: GPT-3.5 Turbo (older v0613)	70.8%	13.7s	$1.50
OpenAI: GPT-4o-mini (2024-07-18)	70.3%	13.2s	$0.38
OpenAI: GPT-4o (2024-08-06)	70.0%	9.5s	$6.25
OpenAI: GPT-4o-mini	67.4%	11.5s	$0.38
OpenAI: GPT-4o-mini Search Preview	66.2%	6.5s	$0.38
OpenAI: GPT-4 Turbo (older v1106)	64.7%	18.1s	$20.00
OpenAI: GPT-4 Turbo Preview	62.6%	15.8s	$20.00
OpenAI: GPT-4	61.7%	12.1s	$45.00
OpenAI: GPT-4 Turbo	61.5%	26.5s	$20.00
OpenAI: GPT-4 (older v0314)	59.7%	16.8s	$45.00
OpenAI: GPT-4.1 Nano	59.2%	5.3s	$0.25
OpenAI: GPT-3.5 Turbo 16k	48.2%	4.2s	$3.50
OpenAI: GPT-3.5 Turbo	42.9%	4.0s	$1.00
OpenAI: GPT-3.5 Turbo Instruct	27.5%	4.4s	$1.75
OpenAI: GPT-4o Audio	—	—	$6.25
OpenAI: gpt-oss-20b	—	—	Free
OpenAI: GPT Audio	—	—	$6.25
OpenAI: GPT Audio Mini	—	—	$1.50
OpenAI: Codex Mini	—	—	$3.75

Benchmark Performance

How this model performs across different benchmarks

Benchmark	Score	Rank
Venture Capital Terms Benchmark	94.4%	16 / 25
Spatial Reasoning: Germany	93.2%	15 / 35
Character Frequency Bench	90.9%	6 / 35
Money Boy Cultural Literacy Test	32.9%	15 / 35
Niederstetten Benchmark	20.0%	32 / 41
German Memelord Bench	9.0%	25 / 35

Price vs Performance

Compare cost efficiency across all models

Current model (baseline)

Other models (relative score)

Y-axis shows score difference from shared benchmarks. X-axis uses log scale.

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