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
Qwen3 ASR 1.7BQwen Remove
Grok 4.7SpaceXAI Remove
MAI-Transcribe 2Microsoft Remove

4 is the maximum. Remove one to add another.

gemini-3.5-transcribe vs qwen3-asr-1.7b vs grok-4.7 vs mai-transcribe-2
Attribute	Gemini 3.5 Transcribegemini-3.5-transcribe	Qwen3 ASR 1.7Bqwen3-asr-1.7b	Grok 4.7grok-4.7	MAI-Transcribe 2mai-transcribe-2
Pricing
Input	$0 / 1M	$0 / 1M	$1.60 / 1M	$0 / 1M
Output	$0 / 1M	$0 / 1M	$4.80 / 1M	$0 / 1M
Cache Write (5m)	Not applicable	Not applicable	$1.60 / 1M	Not applicable
Cache Write (1h)	Not applicable	Not applicable	$1.60 / 1M	Not applicable
Cache Read	Not applicable	Not applicable	$1.60 / 1M	Not applicable
Web Search	$0 / 1M	$0 / 1M	$0 / 1M	$0 / 1M
Context
Max context	98.3K	N/A	500K	N/A
Max output	N/A	N/A	N/A	N/A
Capabilities
Vision	Yes	No	Yes	No
Function Calling	Yes	No	Yes	No
JSON Mode	No	No	Yes	No
Streaming	No	No	Yes	No
Catalogue
Provider	Google	Qwen	SpaceXAI	Microsoft
Category	voice	voice	chat	voice
Charge type	Pay As You Go	Pay As You Go	Pay As You Go	Pay As You Go
Released	2026-09-25	2026-08-13	—	2026-09-03
Description
Summary	Google Gemini 3.5 Transcribe speech-to-text. Billed per input and output token.	Qwen3-ASR 1.7B speech-to-text. Billed per second of audio.	Grok 4.7 is SpaceXAI's flagship model for coding, agentic workflows, and professional knowledge work, succeeding Grok 4.6. It is particularly strong at long-running software engineering, self-verification, and long-context execution, while improving capabilities in document drafting, presentations, and other professional tasks. Trained with extended reinforcement learning focused on multi-hour problems, Grok 4.7 is optimized for sustained, complex task execution and natively supports the Grok Bot harness for conversational workflows. It also introduces an enhanced safeguard stack designed to combine strong jailbreak resistance with low refusal rates for legitimate technical work. Reported benchmark results use xhigh reasoning effort.	Microsoft MAI-Transcribe-2 speech-to-text. Billed per second of audio.