OpenAI
·Released on Aug 5, 2025gpt-oss-120b API Benchmarks, Pricing & Provider Data
Compare gpt-oss-120b with another model
Choose a model to open its comparison page.
LLM
gpt-oss-120b benchmark, API pricing, and provider data cover undefined API provider} other undefined API providers}}, with prices starting at $0.0041/M. gpt-oss-120b free API options are available from undefined provider} other undefined providers}}.
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 par...
Quality
#64of 112
57.0
LMSpeed score
Cost
#22of 186
$0.0041/ 1M · 8:1 in:out
$0.0007 in · $0.0035 out
Category Performance
Observed-capability estimates with uncertainty reported separately from independent benchmark families.
- Coverage
- 7 / 8
- Methodology
- V3.0
#1Math53.7Provisional1/4 Measured dimensions
80% interval: 37.6–69.8
foundational math53.7
aa_math_index · artificial_analysis_math_index
competition mathPrior only
proof frontierPrior only
applied tool mathPrior only
#2Instruction following52.2Provisional1/4 Measured dimensions
80% interval: 36.2–68.2
constraint followingPrior only
structured outputPrior only
novel instruction generalization52.2
ifbench · aa_if_bench
long multiturn instructionPrior only
#3Reasoning49RatedGlobal rank #443/4 Measured dimensions
80% interval: 40.3–57.6
abstract logicPrior only
scientific causal50.8
critpt · critpt / gpqa · gpqa / hle · hle
multistep constraints52.2
mmlu_pro · mmlu_pro
evidence verification43.9
lcr · lcr
#4Coding45.7RatedGlobal rank #343/4 Measured dimensions
80% interval: 36.7–54.8
code generation51.2
livecodebench · livecodebench / scicode · scicode
repository engineering44.8
react_native_bench · react_native_evals
debugging testingPrior only
tooling quality41.3
terminalbench · terminal_bench_hard
#5Knowledge37.9Provisional1/4 Measured dimensions
80% interval: 23.9–51.8
broad knowledge37.9
aa_omniscience · aa_omniscience_index / benchlm_category_knowledge · benchlm_category_knowledge
professional knowledgePrior only
factualityPrior only
retrieval open bookPrior only
#6Agents34.3RatedGlobal rank #614/4 Measured dimensions
80% interval: 28.3–40.3
planning39.5
deep_planning · gert_labs
tool use45.9
tau · tau2_bench
environment execution29.2
enterprise_ops_gym · aa_enterprise_ops_gym / gdpval_aa · benchlm_agentic_gdpval_aa / itbench · aa_itbench
recovery reliability22.4
apex_agents · apex_agents_aa / briefcase · aa_briefcase_elo / harvey_lab · aa_harvey_lab
#7Multilingual25.9Provisional1/4 Measured dimensions
80% interval: 8.8–42.9
cross language understanding25.9
global_mmlu · aa_global_mmlu_lite
multilingual generationPrior only
reasoning transferPrior only
low resource robustnessPrior only
No data:Multimodal
Specifications
Input and output token limits for this model, plus how it ranks on long-context understanding.
INPUT
131.1Ktokens
≈ 157.3 pages of text
OUTPUT
118.0Ktokens
8K128K1M4M
131.1K
Features
Technical Details
- Input
- Output
- Total parameters
- 120B
- Released
- Aug 2025
- Knowledge cutoff
- 2024-06-30
- Tokenizer
- GPT
- Architecture
- text->text
- Moderated
- No
- Supported parameters
- frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_atop_ktop_logprobstop_p
Rankings
Excels at
It's decent at
Falls behind in
Detailed scores
Updated: Sep 13, 2026Overall
undefined metric} other undefined metrics}}
Overall score57.0#64 / 112
Overall
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Speed & latency
undefined metric} other undefined metrics}}
Output speed206.3 tok/s#16 / 80Time to first token0.49 s#9 / 80
Speed & latency
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Output speed
LMSpeed rank#16
Score206.3 tok/s
Sourceartificialanalysis.ai
UpdatedSep 13, 2026
Confidence4
Time to first token
LMSpeed rank#9
Score0.49 s
Sourceartificialanalysis.ai
UpdatedSep 13, 2026
Confidence4
Pricing
undefined metric} other undefined metrics}}
Input price$0.150/M#22 / 186Output price$0.545/M#24 / 186
Pricing
undefined metric} other undefined metrics}}
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Agents
V3.0undefined metric} other undefined metrics}} · Rated
Score34.3#6180% interval28.3–40.34/4 Measured dimensionsAA Agentic Index6.2#60 / 70APEX-Agents-AA3.1#21 / 23
Agents
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Planning & decomposition39.5
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
deep_planning · gert_labs · z -1.38 · q 1.00
Tool use45.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
tau · tau2_bench · z -0.65 · q 1.00
Environment & long-horizon execution29.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
enterprise_ops_gym · aa_enterprise_ops_gym · z -3.00 · q 1.00
gdpval_aa · benchlm_agentic_gdpval_aa · z -1.05 · q 1.00
itbench · aa_itbench · z -3.00 · q 1.00
Recovery & completion reliability22.4
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
apex_agents · apex_agents_aa · z -2.93 · q 1.00
briefcase · aa_briefcase_elo · z -3.00 · q 1.00
harvey_lab · aa_harvey_lab · z -3.00 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
AA EnterpriseOps-GymV3 evidence
LMSpeed rank#19
Score25.5
Sourcebenchlm.ai
UpdatedAug 19, 2026
Confidence1
Coding
V3.0undefined metric} other undefined metrics}} · Rated
Score45.7#3480% interval36.7–54.83/4 Measured dimensionsLiveCodeBench70.7%#29 / 115SciCode34.0%#78 / 89
Coding
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Code generation51.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
livecodebench · livecodebench · z 1.34 · q 1.00
scicode · scicode · z -0.96 · q 1.00
Repository engineering44.8
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
react_native_bench · react_native_evals · z -0.69 · q 1.00
Debugging & testingPrior only
Tool-assisted development & quality41.3
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
terminalbench · terminal_bench_hard · z -1.43 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
LiveCodeBenchV3 evidence
LMSpeed rank#29
Score70.7%
Sourceartificialanalysis.ai
UpdatedSep 13, 2026
Confidence4
SciCodeV3 evidence
LMSpeed rank#78
Score34.0%
Sourceartificialanalysis.ai
UpdatedSep 13, 2026
Confidence4
Terminal-Bench HardV3 evidence
LMSpeed rank#10
Score23.5
Sourcebenchlm.ai
UpdatedSep 13, 2026
Confidence2
Reasoning
V3.0undefined metric} other undefined metrics}} · Rated
Score49#4480% interval40.3–57.63/4 Measured dimensionsMMLU-Pro77.5%#77 / 129GPQA67.2%#149 / 218
Reasoning
V3.0undefined metric} other undefined metrics}} · Rated
Dimensions and evidence
Abstract logicPrior only
Scientific & causal reasoning50.8
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
critpt · critpt · z -0.41 · q 1.00
gpqa · gpqa · z 0.02 · q 1.00
hle · hle · z 0.22 · q 1.00
Multi-step constraints52.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
mmlu_pro · mmlu_pro · z 0.09 · q 1.00
Evidence integration & verification43.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
lcr · lcr · z -1.17 · q 1.00
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
MMLU-ProV3 evidence
LMSpeed rank#77
Score77.5%
Sourceartificialanalysis.ai
UpdatedSep 13, 2026
Confidence4
Knowledge
V3.0undefined metric} other undefined metrics}} · Provisional
Score37.980% interval23.9–51.81/4 Measured dimensionsKnowledge score48.9#72 / 83Artificial Analysis Intelligence Index12.3#105 / 117
Knowledge
V3.0undefined metric} other undefined metrics}} · Provisional
Dimensions and evidence
Broad knowledge37.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
aa_omniscience · aa_omniscience_index · z -1.54 · q 1.00
benchlm_category_knowledge · benchlm_category_knowledge · z -1.24 · q 1.00
Professional knowledgePrior only
FactualityPrior only
Retrieval & open-book usePrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Artificial Analysis Intelligence IndexV3 evidence
LMSpeed rank#105
Score12.3
Sourcebenchlm.ai
UpdatedSep 13, 2026
Confidence2
AA-Omniscience Hallucination Rate
LMSpeed rank#17
Score90.8
Sourcebenchlm.ai
UpdatedSep 13, 2026
Confidence2
Math
V3.0undefined metric} other undefined metrics}} · Provisional
Score53.780% interval37.6–69.81/4 Measured dimensionsAA AIME 202593.4#2 / 2
Math
V3.0undefined metric} other undefined metrics}} · Provisional
Dimensions and evidence
Foundational math53.7
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
aa_math_index · artificial_analysis_math_index · z 0.90 · q 1.00
Competition mathPrior only
Advanced proofsPrior only
Applied & tool-assisted mathPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Multilingual
V3.0undefined metric} other undefined metrics}} · Provisional
Score25.980% interval8.8–42.91/4 Measured dimensionsAA Global-MMLU-Lite82.8#5 / 5
Multilingual
V3.0undefined metric} other undefined metrics}} · Provisional
Dimensions and evidence
Cross-language understanding25.9
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
global_mmlu · aa_global_mmlu_lite · z -3.00 · q 0.63
Multilingual generationPrior only
Reasoning transferPrior only
Low-resource robustnessPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Multimodal
V3.0undefined metric} other undefined metrics}} · No data
80% interval30.8–69.20/4 Measured dimensionsDesign Arena Website980.0#74 / 78
Multimodal
V3.0undefined metric} other undefined metrics}} · No data
Dimensions and evidence
Perception & OCRPrior only
Document & spatial understandingPrior only
Visual reasoningPrior only
Video & grounded actionPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
Instruction following
V3.0undefined metric} other undefined metrics}} · Provisional
Score52.280% interval36.2–68.21/4 Measured dimensionsAA-IFBench69.0#38 / 84Instruction Following score84.3#34 / 52
Instruction following
V3.0undefined metric} other undefined metrics}} · Provisional
Dimensions and evidence
Constraint followingPrior only
Structured outputPrior only
Novel-instruction generalization52.2
undefined family} other undefined families}} · undefined metric} other undefined metrics}}
ifbench · aa_if_bench · z 0.01 · q 1.00
Multi-turn & long instructionsPrior only
Benchmark
LMSpeed rank
Score
Source
Updated
Confidence
OpenRouter endpoints
22 endpointsThird-party OpenRouter endpoint data, shown separately from LMSpeed measurements. Some 30-minute live performance fields only appear after syncing with an OpenRouter API key.
| Provider endpoint | Input | Output | 1d uptime | 30m latency | 30m throughput | Context / output |
|---|---|---|---|---|---|---|
Amazon Bedrock amazon-bedrock/eu-west-1 | $0.150/M | $0.600/M | 100% | — | — | undefined tokens / undefined tokens |
Cerebras cerebras/fp16 | $0.350/M | $0.750/M | 100.0% | — | — | undefined tokens / undefined tokens |
Amazon Bedrock amazon-bedrock | $0.150/M | $0.600/M | 100.0% | — | — | undefined tokens / undefined tokens |
BaseTen baseten/fp4 | $0.100/M | $0.500/M | 100.0% | — | — | undefined tokens / undefined tokens |
DigitalOcean digitalocean | $0.055/M | $0.385/M | 100.0% | — | — | undefined tokens / undefined tokens |
Groq groq | $0.150/M | $0.600/M | 100.0% | — | — | undefined tokens / undefined tokens |
Parasail parasail/fp4 | $0.100/M | $0.750/M | 100.0% | — | — | undefined tokens / undefined tokens |
AkashML akashml/bf16 | $0.030/M | $0.170/M | 100.0% | — | — | undefined tokens / undefined tokens |
DeepInfra deepinfra/turbo | $0.150/M | $0.600/M | 100.0% | — | — | undefined tokens / undefined tokens |
CoreWeave coreweave/fp4 | $0.030/M | $0.170/M | 100.0% | — | — | undefined tokens / undefined tokens |
Google google-vertex/global | $0.090/M | $0.360/M | 99.9% | — | — | undefined tokens / undefined tokens |
Nebius nebius/fp4 | $0.150/M | $0.600/M | 99.8% | — | — | undefined tokens / undefined tokens |
DeepInfra deepinfra/bf16 | $0.037/M | $0.170/M | 99.7% | — | — | undefined tokens / undefined tokens |
DekaLLM dekallm/bf16 | $0.030/M | $0.180/M | 99.1% | — | — | undefined tokens / undefined tokens |
SambaNova sambanova | $0.140/M | $0.950/M | 99.0% | — | — | undefined tokens / undefined tokens |
Mancer 2 mancer/fp8 | $0.055/M | $0.500/M | 98.0% | — | — | undefined tokens / undefined tokens |
Together together | $0.150/M | $0.600/M | 97.7% | — | — | undefined tokens / undefined tokens |
Phala phala | $0.150/M | $0.600/M | 96.1% | — | — | undefined tokens / undefined tokens |
Novita novita/fp4 | $0.050/M | $0.250/M | 95.5% | — | — | undefined tokens / undefined tokens |
SiliconFlow siliconflow/fp8 | $0.150/M | $0.600/M | 95.3% | — | — | undefined tokens / undefined tokens |
DeepInfra deepinfra/fp8 | $0.200/M | $0.950/M | 92.1% | — | — | undefined tokens / undefined tokens |
Mara mara | $0.150/M | $0.750/M | 82.0% | — | — | undefined tokens / undefined tokens |
Pricing Comparison
Compare gpt-oss-120b API pricing across 33 providers. Prices range from $0.0041/M to $77.05/M. 云AI offers the lowest rate at $0.0041/M. 1 provider offers free API credits or a free tier.
| Provider | Health | Model Variant | Group | Input ($/M) | Output ($/M) | Speed (t/s) | First token | Audit |
|---|---|---|---|---|---|---|---|---|
L1 100% | gpt-oss-120b | default | -86%$0.021/M | -85%$0.082/M | — | — | — | |
L1 100% | gpt-oss-120b | default | -91%$0.013/M Cache read$0.013/MCache write$0/MCache write 1h$0/M | -87%$0.072/M | — | — | — | |
L1 100% | openai/gpt-oss-120b | default | $1.00/M Cache read$0.100/M | $2.00/M | — | — | — | |
L1 100% | gpt-oss-120b | Azure-Gpt-2 | -78%$0.034/M | -75%$0.134/M | — | — | — | |
L1 100% | openai/gpt-oss-120b | default | $10.27/M | $10.27/M | — | — | — | |
L1 99% | gpt-oss-120b | Azure-Gpt-2 | -78%$0.034/M | -75%$0.134/M | — | — | — | |
L1 100% | GPT-OSS-120B | default | $75.00/M | $75.00/M | — | — | — | |
S3AI API Free | L1 94% L2 97% | go/gpt-oss-120b | Go Plan | Free | Free | — | — | — |
L1 95% | openai/gpt-oss-120b:free | default | $75.00/M | $75.00/M | — | — | — | |
L1 39% | accounts/fireworks/models/gpt-oss-120b | 测试专用 | $20.55/M Cache read$2.05/M | $82.19/M | — | — | — | |
L1 100% | gpt-oss-120b | default | -50%$0.075/M | -45%$0.301/M | — | — | — | |
L1 100% | laohuang/openai/gpt-oss-120b | default | $0.010/request | - | — | — | — | |
L1 0% | gpt-oss-120b | default | $3.17/M Cache read$3.17/M | $13.86/M | — | — | — |
Alternatives & Similar Models
DeepSeek V4 Flash
deepseek-v4-flash
DeepSeek V4 Flash is a fast, cost-efficient language model in the DeepSeek V4 family, optimized for low-latency chat, coding assistance, and high-throughput API workloads while retaining strong reasoning quality.
15 shared providers
Claude Opus 4.6
claude-opus-4-6
Anthropic Claude Opus 4.6 is the most capable Claude Opus tier, optimized for complex analysis, long-horizon coding, and high-stakes enterprise reasoning workloads.
15 shared providers
GLM-5.1
glm-5-1
Zhipu GLM-5.1 is a next-generation GLM model aimed at frontier reasoning, coding, and bilingual agent applications.
15 shared providers
GPT-5.4
gpt-5-4
OpenAI GPT-5.4 extends the GPT-5 family with stronger instruction following, deeper tool use, and improved performance on coding, math, and long-document analysis.
14 shared providers
GPT-5.2
gpt-5-2
OpenAI GPT-5.2 is a GPT-5 series model emphasizing advanced reasoning, multimodal understanding, and high-quality outputs for complex enterprise workloads.
14 shared providers
GPT-5.4 Mini
gpt-5-4-mini
OpenAI GPT-5.4 Mini is a compact language model in the GPT-5 series, optimized for quick responses and high throughput.
14 shared providers
Frequently Asked Questions
- What benchmark data does gpt-oss-120b include?
- LMSpeed shows gpt-oss-120b benchmark context, API price, output speed, first-token latency, and provider data across 34 providers when those signals are available.
- What is the gpt-oss-120b API price?
- gpt-oss-120b has pricing from undefined provider} other undefined providers}}, ranging from $0.0041/M to $77.05/M. 云AI has the lowest listed price.
- What does the gpt-oss-120b API pricing table include?
- The gpt-oss-120b API pricing table compares 34 providers by input price, output price, free tier status, speed, first-token latency, and recent health data when available.
- Which provider has the cheapest gpt-oss-120b API pricing?
- 云AI currently has the lowest listed gpt-oss-120b price at $0.0041/M across undefined provider} other undefined providers}}.
- Is gpt-oss-120b API free?
- Yes, gpt-oss-120b free API options are available through 1 providerundefined other undefined} on LMSpeed, including S3AI API. These providers offer free API credits or a free tier with no per-token charges.
- Where can I get gpt-oss-120b free API access?
- LMSpeed currently lists 1 free API providerundefined other undefined} for gpt-oss-120b: S3AI API. Check each provider row before using it because free tier limits can change.