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.

Gemini 3.5 TranscribeGoogle Remove
Universal-3.5 ProAssemblyAI Remove
Perceptron Mk1.5Perceptron Remove
MiniMax M2MiniMax Remove

4 is the maximum. Remove one to add another.

gemini-3.5-transcribe vs universal-3-5-pro vs perceptron-mk1.5 vs minimax-m2
Attribute	Gemini 3.5 Transcribegemini-3.5-transcribe	Universal-3.5 Prouniversal-3-5-pro	Perceptron Mk1.5perceptron-mk1.5	MiniMax M2minimax-m2
Pricing
Input	— Not priced per input token	— Not priced per input token	$0.15 / 1M	$0.15 / 1M
Output	— Not priced per output token	— Not priced per output token	$1.50 / 1M	$0.45 / 1M
Cache Write (5m)	Not applicable	Not applicable	$0.15 / 1M	$0.15 / 1M
Cache Write (1h)	Not applicable	Not applicable	$0.15 / 1M	$0.15 / 1M
Cache Read	Not applicable	Not applicable	$0.15 / 1M	$0.15 / 1M
Web Search	$0 / 1M	$0 / 1M	$0 / 1M	$0 / 1M
Context
Max context	98.3K	N/A	36.9K	196.6K
Max output	N/A	N/A	N/A	N/A
Capabilities
Vision	Yes	No	No	Yes
Function Calling	Yes	No	No	Yes
JSON Mode	No	No	No	Yes
Streaming	No	No	Yes	Yes
Catalogue
Provider	Google	AssemblyAI	Perceptron	MiniMax
Category	voice	voice	chat	chat
Charge type	Pay As You Go	Pay As You Go	Pay As You Go	Pay As You Go
Released	2026-09-25	2026-09-22	2026-09-25	—
Description
Summary	Google Gemini 3.5 Transcribe speech-to-text. Billed per input and output token.	AssemblyAI Universal-3.5 Pro speech-to-text. Billed per second of audio.	Perceptron Mk1.5 chat model that accepts audio input. Audio and text input are billed per token.	MiniMax-M2 is a compact, high-efficiency model with 10B active (230B total) parameters, optimized for coding and agentic workflows. It delivers near-frontier reasoning and tool use, excels at multi-file coding tasks and compile-run-fix loops, and performs strongly on benchmarks like SWE-Bench and Terminal-Bench. It also handles long-horizon planning and recovery in agent evaluations, ranking among the top open models across reasoning domains. With fast inference and low cost, it’s ideal for large-scale agents and developer assistants — and works best when reasoning is preserved across turns.