Where is the gpt-oss-120b documentation?

The gpt-oss-120b documentation on OpenRouter includes the LLM Quantization: What fp8 and fp4 Do to Quality and Price blog post. The quickstart covers API keys and your first request, and the API reference lists every parameter.

How much does gpt-oss-120b cost?

gpt-oss-120b costs $0.03/M input tokens and $0.15/M output tokens, with separate rates for Cache Read at $0.015/M tokens.

What is the context length of gpt-oss-120b?

gpt-oss-120b has a 131,072 token context window. It supports up to 131,072 completion tokens.

Does gpt-oss-120b support tool calling and structured outputs?

Yes. gpt-oss-120b accepts tools and tool_choice for function calling on 19 of the 22 providers serving it, and requests that send tools are routed to those providers. It also supports structured outputs via a JSON schema in response_format.

Which providers serve gpt-oss-120b?

gpt-oss-120b is served by 22 providers on OpenRouter: NEAR AI, Venice, CoreWeave, DekaLLM, AkashML, DeepInfra (fp8), Prime Intellect, Crusoe and 14 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

When was gpt-oss-120b released?

gpt-oss-120b was released on August 5, 2025. Its knowledge cutoff is June 30, 2024.

OpenAI: gpt-oss-120b (exacto)

openai/gpt-oss-120b:exacto

Model weights

Compare

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.

Modalities

In / out price

$0.03 / $0.15per 1M

Context

131K

Released

Aug 5, 2025

Knowledge cutoff

Jun 2024

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).

Latency / throughput

Provider	Input /M	Output /M	Cache read /M	Latency	Throughput	Uptime
	$0.05	$0.25	$0.05	0.22s	226 tok/s	100.00%
	$0.03	$0.187	$0.03	0.36s	106 tok/s	99.61%
	$0.15	$0.60	$0.075	0.18s	371 tok/s	99.64%
	$0.06	$0.40	$0.03	0.78s	123 tok/s	100.00%
	$0.03	$0.15	$0.015	0.56s	76 tok/s	99.96%
	$0.055	$0.25	$0.055	0.68s	87 tok/s	99.42%
	$0.05	$0.107	$0.045	1.14s	83 tok/s	99.93%
	$0.06	$0.42	$0.012	0.75s	62 tok/s	100.00%
	$0.14	$0.95	--	1.03s	116 tok/s	99.89%
	$0.03	$0.18	$0.03	0.53s	46 tok/s	97.09%
	$0.03	$0.15	$0.03	0.78s	41 tok/s	98.12%
	$0.15	$0.60	$0.075	1.55s	55 tok/s	97.85%
	$0.037	$0.17	--	0.37s	36 tok/s	99.93%
	$0.25	$0.45	$0.085	1.52s	42 tok/s	86.31%
	$0.20	$0.95	--	1.29s	61 tok/s	90.00%
	$0.10	$0.50	$0.10	0.14s	232 tok/s	99.71%
	$0.10	$0.50	$0.10	0.17s	178 tok/s	99.77%
	$0.15	$0.60	--	0.52s	356 tok/s	96.88%
	$0.15	$0.60	--	0.53s	156 tok/s	97.76%
	$0.10	$0.75	$0.055	0.54s	113 tok/s	99.49%
	$0.15	$0.60	--	0.23s	122 tok/s	99.01%
Fast	$0.15	$0.75	--	0.27s	116 tok/s	99.61%
	$0.15	$0.60	$0.10	0.23s	300 tok/s	99.93%
	$0.35	$0.75	$0.35	0.18s	1,040 tok/s	100.00%
	$0.03	$0.17	$0.03	0.30s	46 tok/s	99.35%
	$0.09	$0.36	--	0.20s	213 tok/s	39.86%

Input $0.05Output $0.25
226 tok/s100.00% uptime
Input $0.03Output $0.187
106 tok/s99.61% uptime
Input $0.15Output $0.60
371 tok/s99.64% uptime
Input $0.06Output $0.40
123 tok/s100.00% uptime
Input $0.03Output $0.15
76 tok/s99.96% uptime
Input $0.055Output $0.25
87 tok/s99.42% uptime
Input $0.05Output $0.107
83 tok/s99.93% uptime
Input $0.06Output $0.42
62 tok/s100.00% uptime
Input $0.14Output $0.95
116 tok/s99.89% uptime
Input $0.03Output $0.18
46 tok/s97.09% uptime
Input $0.03Output $0.15
41 tok/s98.12% uptime
Input $0.15Output $0.60
55 tok/s97.85% uptime
Input $0.037Output $0.17
36 tok/s99.93% uptime
Input $0.25Output $0.45
42 tok/s86.31% uptime
Input $0.20Output $0.95
61 tok/s90.00% uptime
Input $0.10Output $0.50
232 tok/s99.71% uptime
Input $0.10Output $0.50
178 tok/s99.77% uptime
Input $0.15Output $0.60
356 tok/s96.88% uptime
Input $0.15Output $0.60
156 tok/s97.76% uptime
Input $0.10Output $0.75
113 tok/s99.49% uptime
Input $0.15Output $0.60
122 tok/s99.01% uptime
Input $0.15Output $0.75
Fast
116 tok/s99.61% uptime
Input $0.15Output $0.60
300 tok/s99.93% uptime
Input $0.35Output $0.75
1,040 tok/s100.00% uptime
Input $0.03Output $0.17
46 tok/s99.35% uptime
Input $0.09Output $0.36
213 tok/s39.86% uptime

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Weighted average

Input

$0.08012/M tokens

Output

$0.3249/M tokens

gpt-oss-120b — Price history

Price history

No effective price data for this selection

	$0.0211	$0.1490	58.8%	18.6%
	$0.0298	$0.1798	10.7%	1.8%
	$0.0298	$0.1863	59.9%	8.9%
	$0.0298	$0.1488	51.6%	0.8%
	$0.0299	$0.1691	86.8%	15.9%
	$0.0353	$0.0793	48.8%	1.7%
	$0.0368	$0.1694	0.0%	3.6%
	$0.0433	$0.3995	55.4%	1.2%
	$0.0479	$0.4192	24.4%	0.0%
	$0.0498	$0.2492	53.4%	10.9%
	$0.0543	$0.2490	23.8%	0.1%
	$0.0863	$0.7485	30.4%	0.2%
	$0.0998	$0.4995	57.8%	1.5%
	$0.0999	$0.4994	67.5%	0.4%
	$0.1012	$0.5991	64.9%	20.3%
	$0.1123	$0.5989	75.1%	0.2%
	$0.1399	$0.9490	0.0%	0.4%
	$0.1497	$0.5978	0.0%	0.4%
Fast	$0.1497	$0.7489	34.4%	0.3%
	$0.1498	$0.5995	0.0%	0.0%
	$0.1499	$0.5990	47.7%	0.1%
	$0.1500	$0.5996	0.0%	0.8%
	$0.1998	$0.9462	0.0%	0.6%
	$0.2399	$0.4487	6.0%	0.2%
	$0.3498	$0.7490	66.5%	8.7%
	$0.0898	$0.3589	70.3%	2.8%

Effective in $0.0211Effective out $0.1490
Token share
18.6%
Effective in $0.0298Effective out $0.1798
Token share
1.8%
Effective in $0.0298Effective out $0.1863
Token share
8.9%
Effective in $0.0298Effective out $0.1488
Token share
0.8%
Effective in $0.0299Effective out $0.1691
Token share
15.9%
Effective in $0.0353Effective out $0.0793
Token share
1.7%
Effective in $0.0368Effective out $0.1694
Token share
3.6%
Effective in $0.0433Effective out $0.3995
Token share
1.2%
Effective in $0.0479Effective out $0.4192
Token share
0.0%
Effective in $0.0498Effective out $0.2492
Token share
10.9%
Effective in $0.0543Effective out $0.2490
Token share
0.1%
Effective in $0.0863Effective out $0.7485
Token share
0.2%
Effective in $0.0998Effective out $0.4995
Token share
1.5%
Effective in $0.0999Effective out $0.4994
Token share
0.4%
Effective in $0.1012Effective out $0.5991
Token share
20.3%
Effective in $0.1123Effective out $0.5989
Token share
0.2%
Effective in $0.1399Effective out $0.9490
Token share
0.4%
Effective in $0.1497Effective out $0.5978
Token share
0.4%
Effective in $0.1497Effective out $0.7489
FastToken share
0.3%
Effective in $0.1498Effective out $0.5995
Token share
0.0%
Effective in $0.1499Effective out $0.5990
Token share
0.1%
Effective in $0.1500Effective out $0.5996
Token share
0.8%
Effective in $0.1998Effective out $0.9462
Token share
0.6%
Effective in $0.2399Effective out $0.4487
Token share
0.2%
Effective in $0.3498Effective out $0.7490
Token share
8.7%
Effective in $0.0898Effective out $0.3589
Token share
2.8%

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Throughput

1040tok/s

P50, best across providers

Latency

0.14s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchAmazon Bedrock (EU)76.8%--Amazon Bedrock76.4%--Google Vertex74.5%--Parasail (Fast)78.8%68.0%AkashML77.6%66.9%io.net77.2%67.3%Parasail77.1%65.2%

+21 more providers

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Uptime (3d)

100.00%

Availability (3d)

99.13%

Availability over the last 3 days

Last 72 hours

Availability 99.13%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.19%

Without Routing

91.55%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for OpenAI: gpt-oss-120b (Artificial Analysis, Design Arena and Vals AI)
Source	Benchmark	Score
Artificial Analysis	gpt-oss-120b (High) Intelligence Index	3.8
Artificial Analysis	gpt-oss-120b (High) Coding Index	30.4
Artificial Analysis	gpt-oss-120b (High) Agentic Index	3.7
Artificial Analysis	gpt-oss-120b (High) GPQA Diamond	78.2%
Artificial Analysis	gpt-oss-120b (High) HLE	19.6%
Artificial Analysis	gpt-oss-120b (High) IFBench	69.0%
Artificial Analysis	gpt-oss-120b (High) τ²-Bench Telecom	65.8%
Artificial Analysis	gpt-oss-120b (High) AA-LCR	52.0%
Artificial Analysis	gpt-oss-120b (High) τ-Bench Banking	12.8%
Artificial Analysis	gpt-oss-120b (High) GDPval-AA	5.8%
Artificial Analysis	gpt-oss-120b (High) CritPt	1.1%
Artificial Analysis	gpt-oss-120b (High) LiveCodeBench	87.8%
Artificial Analysis	gpt-oss-120b (High) SciCode	34.0%
Artificial Analysis	gpt-oss-120b (High) Terminal-Bench Hard	23.5%
Artificial Analysis	gpt-oss-120b (High) Terminal-Bench 4.0	0.0%
Artificial Analysis	gpt-oss-120b (High) Terminal-Bench 2.1	26.2%
Artificial Analysis	gpt-oss-120b (High) MMLU-Pro	80.8%
Artificial Analysis	gpt-oss-120b (High) AA-Omniscience Accuracy	21.7%
Artificial Analysis	gpt-oss-120b (High) AA-Omniscience Non-Hallucination Rate	13.1%
Artificial Analysis	gpt-oss-120b (High) AIME 2025	93.4%
Artificial Analysis	gpt-oss-120b (Low) Coding Index	21.2
Artificial Analysis	gpt-oss-120b (Low) GPQA Diamond	67.2%
Artificial Analysis	gpt-oss-120b (Low) HLE	5.9%
Artificial Analysis	gpt-oss-120b (Low) IFBench	58.3%
Artificial Analysis	gpt-oss-120b (Low) τ²-Bench Telecom	45.0%
Artificial Analysis	gpt-oss-120b (Low) AA-LCR	46.0%
Artificial Analysis	gpt-oss-120b (Low) τ-Bench Banking	2.9%
Artificial Analysis	gpt-oss-120b (Low) GDPval-AA	0.0%
Artificial Analysis	gpt-oss-120b (Low) CritPt	0.0%
Artificial Analysis	gpt-oss-120b (Low) LiveCodeBench	70.7%
Artificial Analysis	gpt-oss-120b (Low) Terminal-Bench Hard	5.3%
Artificial Analysis	gpt-oss-120b (Low) Terminal-Bench 2.1	13.9%
Artificial Analysis	gpt-oss-120b (Low) MMLU-Pro	77.5%
Artificial Analysis	gpt-oss-120b (Low) AIME 2025	66.7%
Design Arena	GPT OSS 120B Models Arena 3D Elo	881
Design Arena	GPT OSS 120B Models Arena Code Categories Elo	952
Design Arena	GPT OSS 120B Models Arena Data Visualization Elo	984
Design Arena	GPT OSS 120B Models Arena Game Development Elo	981
Design Arena	GPT OSS 120B Models Arena UI Component Elo	917
Design Arena	GPT OSS 120B Models Arena Website Elo	960
Vals AI	GPT OSS 120B AIME	92.6%
Vals AI	GPT OSS 120B CaseLaw v2	48.8%
Vals AI	GPT OSS 120B CorpFin v2	58.2%
Vals AI	GPT OSS 120B Finance Agent v1.1	21.5%
Vals AI	GPT OSS 120B GPQA Diamond	78.5%
Vals AI	GPT OSS 120B LiveCodeBench	83.2%
Vals AI	GPT OSS 120B LegalBench	75.9%
Vals AI	GPT OSS 120B MATH 500	94.8%
Vals AI	GPT OSS 120B MedQA	91.4%
Vals AI	GPT OSS 120B MGSM	92.0%
Vals AI	GPT OSS 120B MMLU Pro	79.2%
Vals AI	GPT OSS 120B SWE-bench	33.6%
Vals AI	GPT OSS 120B TaxEval v2	71.6%
Vals AI	GPT OSS 120B Terminal-Bench 2.0	19.1%

Available through EU in-region routing

OpenAI: gpt-oss-120b (exacto) is available through EU in-region routing. Requests sent to eu.openrouter.ai are decrypted in the EU and routed only to shared provider endpoints in Europe (the EU/EEA, the UK, and Switzerland).

Providers in Europe

Amazon Bedrock

Prices in Europe

Amazon Bedrock: $0.15/M tokens input, $0.60/M tokens output

Base URL

https://eu.openrouter.ai/api/v1

Plan

Requires Business or Enterprise. Self-serve on Business: no contract, no monthly minimum. OpenRouter Business for EU in-region routing

Fails closed: if no shared provider endpoint in Europe is available, the request errors instead of leaving the region. Requests served by your own BYOK key go to your deployment’s region, which you control. In-region routing docs

More models from OpenAI

GPT Image 2.6 Sunburst

GPT Image 2.6 Sunburst is OpenAI's image generation and editing model optimized for detail: dense compositions, text-heavy layouts, and edits that keep details consistent from one edit to the next at medium or high quality, generating about 50% faster than GPT Image 2.5. It takes text and image inputs and produces images, with quality tiers from low to max, transparent backgrounds, and custom sizes in multiples of 16 pixels at aspect ratios from 1:3 to 3:1. GPT Image 2.6 introduces a complexity setting for how much the model plans before rendering, markup-guided edits, animated GIF output, and SVG output. Choose Flare instead when speed matters more than fine detail.

Image$8 / $30

GPT Image 2.6 Flare

GPT Image 2.6 Flare is OpenAI's image generation and editing model optimized for speed: fast iteration and interactive text-to-image and image-to-image workflows, generating about 50% faster than GPT Image 2.5. It takes text and image inputs and produces images, with quality tiers from low to max, transparent backgrounds, and custom sizes in multiples of 16 pixels at aspect ratios from 1:3 to 3:1. GPT Image 2.6 introduces a complexity setting for how much the model plans before rendering, markup-guided edits, animated GIF output, and SVG output. Choose Sunburst instead when detail and edit-to-edit consistency matter more than latency.

Image$8 / $30

GPT-6.1 Terra Pro

GPT-6.1 Terra Pro is the same underlying model as GPT-6.1 Terra, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6.1 Terra and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6.1 Terra instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$1 / $5

GPT-6.1 Terra Pro (batch)

GPT-6.1 Terra Pro is the same underlying model as GPT-6.1 Terra, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$0.50 / $2.50

GPT-6.1 Terra

GPT-6.1 Terra is the mid-tier model in OpenAI's GPT-6.1 series, priced below GPT-6.1 Sol. It is suited for everyday agentic, coding, and writing work, is particularly strong at understanding screenshots and on-screen content, and is efficient enough for high-volume workloads.

Text1.1M context$1 / $5

GPT-6.1 Terra (batch)

Text1.1M context$0.50 / $2.50

GPT-6 Luna Decisions

GPT-6 Luna Decisions is GPT-6 Luna served through OpenAI's Decisions API. Instead of generating text, it reads the content passed as state (text, JSON, or images) and returns typed, probabilistic answers to named questions in a single request: the probability of yes for a yes/no question (noul), a probability for every option plus the most likely one (choice), or a probability for every level of an ordered rubric plus the expected score (score).

It is built for classification, routing, moderation, and rubric grading where application code thresholds the returned numbers rather than parsing a chat reply. A request can carry up to 200 questions about the same content.

Decisions$0.10 / $0

GPT-6.1 Sol Pro

GPT-6.1 Sol Pro is the same underlying model as GPT-6.1 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6.1 Sol and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6.1 Sol instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$2 / $10

GPT-6.1 Sol Pro (batch)

GPT-6.1 Sol Pro is the same underlying model as GPT-6.1 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$1 / $5

GPT-6.1 Sol

GPT-6.1 Sol is an upgrade to GPT-6 Sol from OpenAI, positioned below the flagship GPT-6 Astra in the GPT-6 series. It is suited for agentic coding, computer use, document-heavy professional work, and multi-step business workflow automation, and approaches Astra-level results on these tasks at a much lower cost. Compared with GPT-6 Sol, it makes fewer factual errors and is more reliable at respecting explicit restrictions and user intent during agentic tasks.

Text1.1M context$2 / $10

GPT-6.1 Sol (batch)

Text1.1M context$1 / $5

GPT-6 Luna Pro

GPT-6 Luna Pro is the same underlying model as GPT-6 Luna, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6 Luna and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6 Luna instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$0.10 / $0.50

GPT-6 Luna Pro (batch)

GPT-6 Luna Pro is the same underlying model as GPT-6 Luna, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$0.05 / $0.25

GPT-6 Luna

GPT-6 Luna is the fast, cost-efficient model in OpenAI's GPT-6 series, positioned below GPT-6 Sol. It is suited for high-volume and latency-sensitive workloads such as chat, classification, and lightweight agentic tasks, and at higher reasoning effort it can take on complex software engineering and computer-use tasks that previously called for a Sol-tier model. It shares the GPT-6 family's gains in factual reliability and its clearer, more concise communication style.

Text1.1M context$0.10 / $0.50

GPT-6 Luna (batch)

Text1.1M context$0.05 / $0.25

GPT-6 Sol Pro

GPT-6 Sol Pro is the same underlying model as GPT-6 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Cost note: pro mode spends far more reasoning tokens per request, so a typical request costs several times more than the same request on GPT-6 Sol and takes much longer to complete. It is intended for hard, high-stakes problems where the extra accuracy justifies the cost. For everyday coding, agentic, and chat workloads, use GPT-6 Sol instead.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$2 / $10

GPT-6 Sol Pro (batch)

GPT-6 Sol Pro is the same underlying model as GPT-6 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks.

Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Text1.1M context$1 / $5

GPT-6 Sol

GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional work, agentic coding, business workflow automation, and computer use, and is particularly strong at long-horizon software engineering tasks in real codebases. It approaches Astra-level factual reliability at a much lower cost and shares Astra's clearer, more concise communication style in technical and coding conversations.

Text1.1M context$2 / $10

GPT-6 Sol (batch)

Text1.1M context$1 / $5

GPT Astra Latest

This model always redirects to the latest model in the GPT Astra family.

Text1.1M context

OpenAI: gpt-oss-120b (exacto)

Providers