Compare AI models

Specifications
Parameters
4B—
Context window
128k128k
These models have no shared benchmark scores.
Overview
CompanyNVIDIAOpenAI
Release dateMay 20 2025Feb 12 2026
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Llama Nemotron Nano 4B v1.1 and GPT-5.3-Codex-Spark don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Llama Nemotron Nano 4B v1.1 shipped 268 days before GPT-5.3-Codex-Spark, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Llama Nemotron Nano 4B v1.1) vs 128k (GPT-5.3-Codex-Spark). Llama Nemotron Nano 4B v1.1 is open weight, while GPT-5.3-Codex-Spark is closed.

Direct benchmark comparisons are unavailable — Llama Nemotron Nano 4B v1.1 and GPT-5.3-Codex-Spark don't publish scores on any of the same benchmarks.