Compare AI models

Specifications
Parameters
—4B
Context window
1M256k
API pricing
Input price
$2.00—
Output price
$12.00—
Cached input price
$0.20—
Cheapest input
$1.00Google AI Studio—
Cheapest output
$6.00Google AI Studio—
$1,000Google list price—
Benchmarks
τ²-bench · retail
90.8%39.8%
τ²-bench · telecom
99.3%33%
GPQA Diamond
94.3%53.2%
Overview
CompanyGoogleNVIDIA
Release dateFeb 19 2026Mar 16 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Gemini 3.1 Pro leads Nemotron 3 Nano 4B on 3 of the 3 benchmarks they both report (τ²-bench, GPQA Diamond). Only Gemini 3.1 Pro has a verified first-party price for an available API: $2.00 per million input tokens and $12.00 per million output tokens. No currently available pay-as-you-go API rate is tracked for Nemotron 3 Nano 4B. Gemini 3.1 Pro shipped 25 days before Nemotron 3 Nano 4B, so benchmark comparisons should account for the intervening progress.

Context windows are 1M (Gemini 3.1 Pro) vs 256k (Nemotron 3 Nano 4B). Gemini 3.1 Pro is closed, while Nemotron 3 Nano 4B is open weight.

On τ²-bench · retail, Gemini 3.1 Pro leads at 90.8% vs Nemotron 3 Nano 4B at 39.8%. On τ²-bench · telecom, Gemini 3.1 Pro leads at 99.3% vs Nemotron 3 Nano 4B at 33%. On GPQA Diamond, Gemini 3.1 Pro leads at 94.3% vs Nemotron 3 Nano 4B at 53.2%.