openai /

gpt-oss-120b-Turbo

52.5 DZD in 210 DZD out/ 1M tokens

The gpt‑oss‑120b‑Turbo is an open‑weight large language model from OpenAI’s gpt‑oss series. It uses a Mixture‑of‑Experts Transformer architecture with about 117 billion parameters, but only ~5.1 billion are active per pass, making it faster and more efficient. The Turbo version is optimized for lower latency, higher throughput, configurable reasoning, and agentic tasks like function calling and tool use. It runs on a single H100 GPU with MXFP4 quantization and is released under the Apache 2.0 license for both research and enterprise use.