Compare AI models

Specifications
Parameters
—27B
Context window
128k256k
These models have no shared benchmark scores.
Overview
CompanyOpenAIQwen
Release dateFeb 12 2026Apr 22 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

GPT-5.3-Codex-Spark and Qwen3.6-27B don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. GPT-5.3-Codex-Spark shipped 69 days before Qwen3.6-27B, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (GPT-5.3-Codex-Spark) vs 256k (Qwen3.6-27B). GPT-5.3-Codex-Spark is closed, while Qwen3.6-27B is open weight.

Direct benchmark comparisons are unavailable — GPT-5.3-Codex-Spark and Qwen3.6-27B don't publish scores on any of the same benchmarks.