by OpenAIRank #162Score 68.5
gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.
| Signal | Normalized | Weight | Contribution | Freshness |
|---|---|---|---|---|
Capabilities capability | 66.7 | 20% | 13.3 | 2026-03-23T03:01:53.837Z |
Benchmarks benchmark | 55.7 | 30% | 16.7 | 2026-03-23T03:01:53.837Z |
Pricing pricing_tier | 0.1 | 15% | 0.0 | 2026-03-23T03:01:53.837Z |
Context Window context_window | 81.2 | 10% | 8.1 | 2026-03-23T03:01:53.837Z |
Recency recency | 91.0 | 15% | 13.7 | 2026-03-23T03:01:53.837Z |
Output Capacity output_capacity | 85.3 | 10% | 8.5 | 2026-03-23T03:01:53.837Z |
| Capability | Supported |
|---|---|
| Vision | No |
| Reasoning | Yes |
| JSON Mode | Yes |
| Streaming | Yes |
| Function Calling | Yes |
| Web Search | No |