Skip to content
Alpha
  • Models
  • Rankings
  • Ori
Talk to sales

Footer

OpenRouterOpenRouter
© 2026 OpenRouter, Inc

Product

  • Chat
  • Rankings
  • Benchmarks
  • Apps
  • Discover
  • Models
  • Ori
  • Collections
  • Providers
  • Tools
  • Pricing
  • Business
  • Enterprise
  • Labs

Company

  • About
  • Blog
  • Customers
  • Careers
    Hiring
  • Privacy
  • Terms of Service
  • Trust Center
  • Support
  • Works With OR
  • Data
  • Brand

Developer

  • Documentation
  • API Reference
  • Developer Platform
  • Status
  • AI Site Map

Connect

  • Discord
  • GitHub
  • LinkedIn
  • X
  • YouTube
Favicon for openai

OpenAI: gpt-oss-120b

openai/gpt-oss-120b

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
Compare
ProvidersPricingPerformanceUptimeBenchmarksAppsActivityFAQExplore

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).

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.

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).

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.

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)
SourceBenchmarkScore
Artificial Analysisgpt-oss-120b (High) Coding Index30.4
Artificial Analysisgpt-oss-120b (High) Agentic Index3.7
Artificial Analysisgpt-oss-120b (High) GPQA Diamond78.2%
Artificial Analysisgpt-oss-120b (High) HLE19.6%
Artificial Analysisgpt-oss-120b (High) IFBench69.0%
Artificial Analysisgpt-oss-120b (High) τ²-Bench Telecom65.8%
Artificial Analysisgpt-oss-120b (High) AA-LCR52.0%
Artificial Analysisgpt-oss-120b (High) τ-Bench Banking12.8%
Artificial Analysisgpt-oss-120b (High) GDPval-AA5.8%
Artificial Analysisgpt-oss-120b (High) CritPt1.1%
Artificial Analysisgpt-oss-120b (High) LiveCodeBench87.8%
Artificial Analysisgpt-oss-120b (High) SciCode34.0%
Artificial Analysisgpt-oss-120b (High) Terminal-Bench Hard23.5%
Artificial Analysisgpt-oss-120b (High) Terminal-Bench 4.00.0%
Artificial Analysisgpt-oss-120b (High) Terminal-Bench 2.126.2%
Artificial Analysisgpt-oss-120b (High) MMLU-Pro80.8%
Artificial Analysisgpt-oss-120b (High) AIME 202593.4%
Artificial Analysisgpt-oss-120b (Low) Coding Index21.2
Artificial Analysisgpt-oss-120b (Low) GPQA Diamond67.2%
Artificial Analysisgpt-oss-120b (Low) HLE5.9%
Artificial Analysisgpt-oss-120b (Low) IFBench58.3%
Artificial Analysisgpt-oss-120b (Low) τ²-Bench Telecom45.0%
Artificial Analysisgpt-oss-120b (Low) AA-LCR46.0%
Artificial Analysisgpt-oss-120b (Low) τ-Bench Banking2.9%
Artificial Analysisgpt-oss-120b (Low) GDPval-AA0.0%
Artificial Analysisgpt-oss-120b (Low) CritPt0.0%
Artificial Analysisgpt-oss-120b (Low) LiveCodeBench70.7%
Artificial Analysisgpt-oss-120b (Low) Terminal-Bench Hard5.3%
Artificial Analysisgpt-oss-120b (Low) Terminal-Bench 2.113.9%
Artificial Analysisgpt-oss-120b (Low) MMLU-Pro77.5%
Artificial Analysisgpt-oss-120b (Low) AIME 202566.7%
Design ArenaGPT OSS 120B Models Arena 3D Elo898
Design ArenaGPT OSS 120B Models Arena Code Categories Elo961
Design ArenaGPT OSS 120B Models Arena Data Visualization Elo987
Design ArenaGPT OSS 120B Models Arena Game Development Elo992
Design ArenaGPT OSS 120B Models Arena UI Component Elo921
Design ArenaGPT OSS 120B Models Arena Website Elo966
Vals AIGPT OSS 120B AIME92.6%
Vals AIGPT OSS 120B CaseLaw v248.8%
Vals AIGPT OSS 120B CorpFin v258.2%
Vals AIGPT OSS 120B Finance Agent v1.121.5%
Vals AIGPT OSS 120B GPQA Diamond78.5%
Vals AIGPT OSS 120B LiveCodeBench83.2%
Vals AIGPT OSS 120B LegalBench75.9%
Vals AIGPT OSS 120B MATH 50094.8%
Vals AIGPT OSS 120B MedQA91.4%
Vals AIGPT OSS 120B MGSM92.0%
Vals AIGPT OSS 120B MMLU Pro79.2%
Vals AIGPT OSS 120B SWE-bench33.6%
Vals AIGPT OSS 120B TaxEval v271.6%
Vals AIGPT OSS 120B Terminal-Bench 2.019.1%

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

Activity

Token volume and request traffic to this model over time.

API

Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.

Available through EU in-region routing

OpenAI: gpt-oss-120b 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

Explore more models

AI Model RankingsRanking
Latency / throughput
$0.030$0.150$0.0150.79s38 tps
100.00%
$0.030$0.150$0.0301.79s32 tps
99.75%
$0.030$0.170$0.0300.33s48 tps
99.97%
$0.030$0.180$0.0300.51s38 tps
99.89%
$0.030$0.187$0.0300.44s107 tps
100.00%
$0.037$0.170--0.47s32 tps
95.72%
$0.050$0.250$0.0500.19s271 tps
100.00%
$0.050$0.250--0.88s42 tps
99.67%
$0.050$0.250--0.70s141 tps
99.78%
$0.060$0.400$0.0300.79s124 tps
100.00%
$0.060$0.420$0.0120.62s66 tps
100.00%
$0.100$0.500$0.1000.14s212 tps
99.34%
$0.100$0.500$0.1000.20s210 tps
98.84%
$0.100$0.750$0.0550.34s222 tps
99.94%
$0.140$0.950--1.24s388 tps
99.94%
$0.150$0.600$0.1000.26s299 tps
100.00%
$0.150$0.600--0.55s302 tps
100.00%
$0.150$0.600$0.0751.12s39 tps
92.30%
$0.150$0.600--0.86s147 tps
99.61%
$0.150$0.600--0.28s74 tps
98.97%
$0.150$0.600$0.0750.16s346 tps
99.94%
$0.150$0.750--1.04s198 tps
97.56%
$0.350$0.750$0.3500.19s858 tps
100.00%
$0.090$0.360--0.88s34 tps
1.39%
$0.200$0.950--6.74s60 tps
27.78%
Not used in Standard routing:Why these endpoints are not used
$0.150$0.600--0.47s144 tps
100.00%
Throughput
858tok/s

P50, best across providers

Latency
0.14s

P50, best provider

AutoExacto benchmarks
GPQA DiamondTAU-BenchAmazon Bedrock77.2%--Amazon Bedrock (EU)76.9%--Sail Research76.6%--DekaLLM77.7%66.1%AkashML77.2%66.1%DeepInfra (bf16)76.3%66.0%DeepInfra65.9%67.2%
+19 more providers
Uptime (3d)The model was reachable. Request routed to a provider.
100.00%
Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.
99.64%

Availability over the last 3 days

Last 72 hours
Availability 99.64%
3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability
99.78%
Without Routing
86.34%

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.

1.
Favicon for https://mira.tg/
Mira
Mira is a Telegram-native AI assistant.
47.7Btokens
2.
Favicon for https://convosight.com/
cvi-magic-tasks-ai-tagging
new
43.9Btokens
3.
Favicon for https://op.gg/
OP.GG
new
29.7Btokens
4.
Favicon for https://apollo.halan.io/
LibreChat
new
14Btokens
5.
Favicon for https://craftrpgs.com/?utm_source=openrouter
Craft
Build and play AI RPGs
12.4Btokens

Frequently asked questions

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.

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.

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

Yes. gpt-oss-120b accepts tools and tool_choice for function calling on 19 of the 23 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.

gpt-oss-120b is served by 23 providers on OpenRouter: NEAR AI, Venice, CoreWeave, DekaLLM, AkashML, DeepInfra (fp8), Crusoe, Mancer and 15 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.

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

More models from OpenAI

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.

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$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)

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$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.

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$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)

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$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.

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$0.05 / $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)

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$0.05 / $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.

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$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)

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$1 / $5
GPT Astra Latest

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

Text1.1M context
GPT Sol Latest

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

Text1.1M context
GPT Terra Latest

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

Text1.1M context