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.

Qwen3 ASR 1.7BQwen Remove
Muse Spark 1.3Meta Remove
Perceptron Mk1.5Perceptron Remove
Grok 4.7SpaceXAI Remove

4 is the maximum. Remove one to add another.

qwen3-asr-1.7b vs muse-spark-1.3 vs perceptron-mk1.5 vs grok-4.7
Attribute	Qwen3 ASR 1.7Bqwen3-asr-1.7b	Muse Spark 1.3muse-spark-1.3	Perceptron Mk1.5perceptron-mk1.5	Grok 4.7grok-4.7
Pricing
Input	$0 / 1M	$1.25 / 1M	$0.15 / 1M	$1.60 / 1M
Output	$0 / 1M	$4.25 / 1M	$1.50 / 1M	$4.80 / 1M
Cache Write (5m)	Not applicable	$1.25 / 1M	$0.15 / 1M	$1.60 / 1M
Cache Write (1h)	Not applicable	$1.25 / 1M	$0.15 / 1M	$1.60 / 1M
Cache Read	Not applicable	$1.25 / 1M	$0.15 / 1M	$1.60 / 1M
Web Search	$0 / 1M	$0 / 1M	$0 / 1M	$0 / 1M
Context
Max context	N/A	1M	36.9K	500K
Max output	N/A	N/A	N/A	N/A
Capabilities
Vision	No	Yes	No	Yes
Function Calling	No	Yes	No	Yes
JSON Mode	No	Yes	No	Yes
Streaming	No	Yes	Yes	Yes
Catalogue
Provider	Qwen	Meta	Perceptron	SpaceXAI
Category	voice	chat	chat	chat
Charge type	Pay As You Go	Pay As You Go	Pay As You Go	Pay As You Go
Released	2026-08-13	—	2026-09-25	—
Description
Summary	Qwen3-ASR 1.7B speech-to-text. Billed per second of audio.	Muse Spark 1.3 is Meta's multimodal reasoning model designed for long-running agentic, multi-agent, and coding workflows. It maintains context and information across extended tasks, enabling reliable execution in complex, multi-step environments. The model is optimized to resolve conflicting information, seek clarification or confirmation when necessary, and execute concisely, making it well suited for autonomous agents, collaborative multi-agent systems, and long-horizon software engineering workflows.	Perceptron Mk1.5 chat model that accepts audio input. Audio and text input are billed per token.	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.