← Back to all models
G

GPT-OSS 120B

by OpenAI
Flagship Free & Open Source 🏆 Ranked #24 of 85
80.9
Overall Score
out of 100
About

OpenAI's 120B open-weight language model, making frontier-class performance available for self-hosted and on-premise deployments. The largest member of OpenAI's open-weights family.

Key Metrics
Context Window
128K
tokens
Avg Response
1300
milliseconds
Input Cost
$0.8
per million tokens
Output Cost
$2.4
per million tokens
Arena ELO
1285
Chatbot Arena rating
MT-Bench
9.0
out of 10
Benchmark Scores
MMLU
87.0%
HumanEval
90.0%
MATH
84.0%
GPQA
58.0%
MT-Bench
90.0/10
Capability Profile
Strengths & Limitations
Strengths
✓ OpenAI quality open-source ✓ Large parameter count ✓ Self-hostable ✓ Fine-tuneable ✓ Strong general capability
Limitations
⚠ Server-scale hardware required ⚠ OpenAI ecosystem dependency ⚠ Large download
Ideal Use Cases
On-premise OpenAI deployment Enterprise AI Research Privacy-sensitive tasks Fine-tuning
Model Details
Provider OpenAI
Released 2025-04-14
Type Free & Open Source
Multimodal No
Tier Flagship
Global rank #24 / 85
Pricing (USD)
Input tokens $0.8/M
Output tokens $2.4/M
Per 1,000 tokens ≈ $0.0008 input / $0.0024 output
All Benchmarks
MMLU 87.0%
HumanEval 90.0%
MATH 84.0%
GPQA 58.0%
MT-Bench 9.0/10
Arena ELO 1285
Compare this model View Rankings

You might also consider