Skip to content

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. S2 ProFish AudioRemove
  2. GPT-6 Luna ProOpenAIRemove
  3. Aura-2DeepgramRemove
  4. Muse Spark 1.3MetaRemove

4 is the maximum. Remove one to add another.

s2-pro vs gpt-6-luna-pro vs aura-2 vs muse-spark-1.3
AttributeS2 Pros2-proGPT-6 Luna Progpt-6-luna-proAura-2aura-2Muse Spark 1.3muse-spark-1.3
Pricing
Input$0 / 1M$0.10 / 1M$0 / 1M$1.25 / 1M
Output$0 / 1M$0.50 / 1M$0 / 1M$4.25 / 1M
Cache Write (5m)Not applicable$0.10 / 1MNot applicable$1.25 / 1M
Cache Write (1h)Not applicable$0.10 / 1MNot applicable$1.25 / 1M
Cache ReadNot applicable$0.10 / 1MNot applicable$1.25 / 1M
Web Search$0 / 1M$0 / 1M$0 / 1M$0 / 1M
Context
Max contextN/A1.1MN/A1M
Max outputN/AN/AN/AN/A
Capabilities
VisionNoYesNoYes
Function CallingNoYesNoYes
JSON ModeNoYesNoYes
StreamingNoYesNoYes
Catalogue
ProviderFish AudioOpenAIDeepgramMeta
Categoryvoicechatvoicechat
Charge typePay As You GoPay As You GoPay As You GoPay As You Go
Released2026-07-29—2026-07-16—
Description
SummaryFish Audio S2 Pro text-to-speech. Billed per UTF-8 byte of input text.GPT-6 Luna Pro uses the same underlying model as GPT-6 Luna, but runs with reasoning.mode set to pro for higher-quality responses on complex tasks. It combines Luna's speed and cost efficiency with deeper reasoning and more reliable multi-step execution, making it well suited for advanced coding, computer use, and agentic workflows where higher solution quality is needed without moving to a larger GPT-6 tier.Deepgram Aura-2 text-to-speech for real-time voice applications. Billed per input character.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.