Skip to content
Apertis

Compare models

Put up to 4 models beside each other — token prices, context windows, capabilities and provider, from the same catalogue the model pages read.

  1. Perceptron Mk1.5PerceptronRemove
  2. Gemini 3.5 TranscribeGoogleRemove
  3. Nemotron 3 Super (Free)NVIDIARemove
perceptron-mk1.5 vs gemini-3.5-transcribe vs nemotron-3-super-120b-a12b:free
AttributePerceptron Mk1.5perceptron-mk1.5Gemini 3.5 Transcribegemini-3.5-transcribeNemotron 3 Super (Free)nemotron-3-super-120b-a12b:free
Pricing
Input$0.15 / 1M— Not priced per input token$0 / 1M
Output$1.50 / 1M— Not priced per output token$0 / 1M
Cache Write (5m)$0.15 / 1MNot applicable—
Cache Write (1h)$0.15 / 1MNot applicable—
Cache Read$0.15 / 1MNot applicable$0 / 1M
Web Search$0 / 1M$0 / 1M—
Cache Write——$0 / 1M
Context
Max context36.9K98.3K262.1K
Max outputN/AN/AN/A
Capabilities
VisionNoYesYes
Function CallingNoYesYes
JSON ModeNoNoYes
StreamingYesNoYes
Catalogue
ProviderPerceptronGoogleNVIDIA
Categorychatvoicechat
Charge typePay As You GoPay As You GoFree
Released2026-09-252026-09-25—
Description
SummaryPerceptron Mk1.5 chat model that accepts audio input. Audio and text input are billed per token.Google Gemini 3.5 Transcribe speech-to-text. Billed per input and output token.NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid Mixture-of-Experts model designed for complex multi-agent and long-horizon reasoning workflows. It activates only 12B parameters per token, enabling high compute efficiency while maintaining strong accuracy on advanced tasks. Built on a hybrid Mamba–Transformer MoE architecture with multi-token prediction (MTP), the model delivers significantly higher token generation throughput than leading open models. It supports a 1M-token context window for long-context reasoning, cross-document analysis, and multi-step task planning. Trained with multi-environment reinforcement learning across diverse benchmarks—including AIME 2025, TerminalBench, and SWE-Bench Verified—Nemotron 3 Super achieves strong performance across reasoning and coding tasks. Released fully open with weights, datasets, and training recipes, it supports flexible customization and secure deployment from local workstations to cloud environments.