Model Overview & Capabilities
gpt-oss-120b (high) is a frontier artificial intelligence model engineered by OpenAI, officially verified in January 2026 and evaluated on the Artificial Analysis Intelligence Index v4.3 with an overall score of 12.0. Architected as a open-weights architecture, the system operates across a context window of 131k tokens, allowing enterprise developer agents and autonomous code evaluation tools to ingest comprehensive repository structures, cross-language modules, and multi-layered documentation hierarchies without degradation of semantic coherence.
Under standardized software engineering benchmarks—including SWE-bench Verified, Terminal-Bench 2.0, Aider Polyglot, and LiveCodeBench—gpt-oss-120b (high) exhibits rigorous multi-step problem solving, deterministic SEARCH/REPLACE diff compliance, and robust tool-use navigation in sandboxed execution environments. Its reasoning capabilities are calibrated to minimize hallucination rates while handling complex syntax transformations, boundary conditions, and continuous integration workflows.
Economically and operationally, gpt-oss-120b (high) generates output tokens at a median throughput of 181 tokens per second, with an observed initial latency of 850 milliseconds. Artificial Analysis rates its estimated cost per task at $0.11 USD, backed by an API token pricing structure of $0.12 per million input tokens and $0.36 per million output tokens (resulting in an effective blended rate of $0.18 per million tokens).