Compare AI models

Specifications
Parameters
—4B
Context window
1M256k
API pricing
Cheapest input
$0.03Particle AI—
Cheapest output
$0.2352DeepInfra—
$25.00Pearl Research · fp8—
These models have no shared benchmark scores.
Overview
CompanyDeepSeekNVIDIA
Release dateOct 17 2026Mar 16 2026
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek-V4.1-Flash-1017 and Nemotron 3 Nano 4B don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Nemotron 3 Nano 4B shipped 215 days before DeepSeek-V4.1-Flash-1017, so benchmark comparisons should account for the intervening progress.

Context windows are 1M (DeepSeek-V4.1-Flash-1017) vs 256k (Nemotron 3 Nano 4B).

Direct benchmark comparisons are unavailable — DeepSeek-V4.1-Flash-1017 and Nemotron 3 Nano 4B don't publish scores on any of the same benchmarks.