OpenAI
OpenAI GPT-6 model for focused, high-volume workloads where cost efficiency is the priority.
1.05M ctxUSD 0.10 / 1M input tok; USD 0.50 / 1M output tok
Introducing gpt-oss-120bOpenAIReleased September 4, 2025
gpt-oss-120b
OpenAI’s premier open-weight model licensed under Apache 2.0, engineered with 117B parameters (5.1B active) to fit on a single H100 GPU.
reasoning-llmOpen WeightsFree Open Weights (Apache 2.0)Context: 131.072K (131,072 tokens)Arena ELO: 1350
Technical Specifications
Architecture Type
Open-Weights Sparse MoE Reasoning Model
Total Parameters
117B
Active Parameters (MoE)
5.1B
Context Window
131.072K (131,072 tokens)
Max Output Tokens
32.768K (32,768 tokens)
Knowledge Cutoff
October 2024
Supported Modalities
text, code, reasoning
License & Access
Apache 2.0
Weights Formats
safetensors, gguf, fp8
Benchmark Evaluations
Math500
88.4
Mmlu Pro
74.2
Aime 2024
76.5
Humaneval
91
Chatbot Arena ELO
1350
Deep Architectural Overview
gpt-oss-120b is OpenAI’s landmark open-weights release. Released under the permissive Apache 2.0 license, it features full chain-of-thought visibility, configurable reasoning effort, native function calling, and fits entirely in 80GB VRAM.
Strengths & Considerations
Core Strengths
- Permissive Apache 2.0 commercial license
- Runs on a single NVIDIA H100 GPU (117B total / 5.1B active)
- Full chain-of-thought transparency and parameter fine-tunability
Known Limitations
- Requires enterprise 80GB VRAM server to run unquantized
Token & API Pricing
Input Tokens (1M)Free (Open Weights)
Output Tokens (1M)Free (Open Weights)
Self-Hosted Min VRAM80 GB
Pricing is verified directly against OpenAI's developer documentation and API rate sheets.
Similar & Alternative Models
Explore other frontier models from OpenAI and comparable reasoning engines.
OpenAI
OpenAI GPT-6 model built for complex coding and agentic workflows, balancing intelligence and cost.
1.05M ctxUSD 2 / 1M input tok; USD 10 / 1M output tok
OpenAI
OpenAI’s premier sixth-generation frontier model, built for the most demanding end-to-end software engineering and scientific research tasks.
1.05M ctx$10.00 / 1M tok (GPT-6 Frontier)
OpenAI
Flagship model family of mid-2026 featuring Sol (flagship), Terra (balanced), Luna (high-speed), and Cyber (authorized vulnerability research).
1.05M ctx$4.00 / 1M tok (Flagship Sol)