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. Muse Spark OpenMetaRemove
  2. Qwen3.8 2.4T A95BAlibabaRemove
  3. Perceptron Mk1.5PerceptronRemove
muse-spark-open vs qwen3.8-2.4t-a95b vs perceptron-mk1.5
AttributeMuse Spark Openmuse-spark-openQwen3.8 2.4T A95Bqwen3.8-2.4t-a95bPerceptron Mk1.5perceptron-mk1.5
Pricing
Input$1.25 / 1M$1.80 / 1M$0.15 / 1M
Output$4.25 / 1M$5.40 / 1M$1.50 / 1M
Cache Write (5m)$1.25 / 1M$1.80 / 1M$0.15 / 1M
Cache Write (1h)$1.25 / 1M$1.80 / 1M$0.15 / 1M
Cache Read$1.25 / 1M$1.80 / 1M$0.15 / 1M
Web Search$0 / 1M$0 / 1M$0 / 1M
Context
Max context1M262K36.9K
Max outputN/AN/AN/A
Capabilities
VisionYesYesNo
Function CallingYesYesNo
JSON ModeYesYesNo
StreamingYesYesYes
Catalogue
ProviderMetaAlibabaPerceptron
Categorychatchatchat
Charge typePay As You GoPay As You GoPay As You Go
Released——2026-09-25
Description
SummaryMuse Spark Open is an open-weight, coding-optimized model designed for agentic software engineering workflows. It delivers strong first-attempt accuracy, reliable tool calling, and deep codebase understanding, enabling effective development across real-world repositories. Optimized for autonomous coding agents, complex code navigation, and tool-driven development, Muse Spark Open is well suited for software engineering automation, debugging, code generation, and multi-step coding workflows where accuracy and reliable execution are essential.Qwen3.8 2.4T A95B is Qwen's open-weight sparse Mixture-of-Experts (MoE) model and the open-weight counterpart to Qwen3.8 Max. It features 2.4T total parameters with 95B activated per token, combining frontier-scale capacity with efficient sparse inference. Designed for coding, research, complex reasoning, and agentic workflows, the model is well suited for demanding long-horizon tasks and advanced autonomous systems while providing the flexibility and customization benefits of open weights.Perceptron Mk1.5 chat model that accepts audio input. Audio and text input are billed per token.