DataAIHub
DataAIHubNews · Research · Tools · Learning
MetaProprietaryCodingReasoningMultimodal

Muse Spark

Meta’s closed Muse Spark 1.3 multimodal reasoning model on Meta Model API and Muse Code.

Meta Superintelligence Labs’ Muse Spark 1.3 — closed multimodal reasoning model for agentic tasks, long-horizon coding, computer use, and 1M-context workflows via Muse Code and the Meta Model API. Succeeds Spark 1.2; max reasoning is still withheld for safety testing.

Why Muse Spark matters

Muse Spark is Meta’s paid API path for agentic coding, computer use, and multimodal agents—distinct from open-weight Llama and from Muse Glimmer (Apache-2.0 30B on-device). Spark 1.3 (2026-09-02) improves long-horizon collaboration; max reasoning is still withheld for safety testing, and open weights are unreleased.

Vision · Audio · Tool calling · Thinking · MCP · Coding

Last reviewed: 4 September 2026

Official pricing →

When to choose Muse Spark

Decision guidance for architects—not a feature list.

Best for

  • Long-horizon coding with Muse Code
  • Personal agentic orchestration
  • Computer-use workflows
  • Multimodal perception + action loops

Avoid if

  • You require open weights or on-prem-only inference
  • You need a mature terminal-agent ecosystem beyond Muse Code beta

Strengths

Qualitative snapshot for architects—not a public ranking.

  • Agentic coding / tools★★★★★
  • Multimodal + computer use★★★★★
  • API price competitiveness★★★★☆
  • Open weights★☆☆☆☆
  • Ecosystem maturity★★★☆☆

Ecosystem

Built by

  • Meta

    Muse Spark is Meta Superintelligence Labs’ paid multimodal agentic model.

Competes with

  • GPT-5.6

    Competes on agentic coding and multimodal tool use via a paid API.

  • Claude Opus

    Competes for long-running agents and computer-use style workflows.

Works with

  • Muse Code

    Terminal coding agent co-trained with Muse Spark 1.3 for long-horizon repo work.

  • Cursor

    Early partners use Muse Spark inside agentic coding harnesses.

  • LangGraph

    Stateful agents can call Muse Spark via OpenAI-compatible Meta Model API.

Recommended for

  • AI Agents

    Built for multi-agent orchestration, tools/MCP, and computer use.

  • Llama Models

    Clarify open Llama vs closed Muse Spark when evaluating Meta stacks.

Often paired with

  • Llama 4

    Llama stays Meta’s open-weight generalist path; Muse Spark is the closed API path.

  • Muse Glimmer

    Glimmer is the Apache-2.0 on-device Muse sibling; Spark is the paid 1M-context API path.

How Muse Spark evolved

Key moments in chronological order.

  1. Model

    Muse Spark 1.3

    Improves long-horizon agentic work and coding in Muse Code and Meta Model API. Asks clarifying questions, confirms consequential actions, and uses fewer turns/tokens than 1.2 per Meta. Previously available reasoning modes are live; max reasoning is still withheld for safety testing. Open weights not released.

  2. Open source

    Muse Glimmer 30B open weights

    Apache-2.0 30B on-device agent model (sibling to closed Muse Spark); Hugging Face meta-models/Muse-Glimmer-30B.

  3. Model

    Muse Spark 1.2

    Coding-focused upgrade with more coding compute and environment diversity; available via Meta Model API with expanded global access.

  4. Product

    Muse Code beta

    Terminal coding agent powered by Muse Spark 1.2 with persistent background subagents and a replay-exact event-log runtime.

  5. Model

    Muse Spark 1.1

    Major gains in agentic coding, tool/computer use, and multimodal understanding.

  6. API

    Meta Model API public preview

    Meta’s first paid model API (OpenAI-compatible) opens Muse Spark to developers.

  7. Product

    Thinking mode in Meta AI

    Consumer Meta AI / meta.ai surfaces expose Muse Spark Thinking mode.

  8. Model

    Muse Spark first release

    Meta Superintelligence Labs ships its first closed multimodal Muse model.

Overview

Meta Superintelligence Labs’ Muse Spark 1.3 — closed multimodal reasoning model for agentic tasks, long-horizon coding, computer use, and 1M-context workflows via Muse Code and the Meta Model API. Succeeds Spark 1.2; max reasoning is still withheld for safety testing.

Capabilities

  • Vision: Yes
  • Audio: Yes
  • Tool calling: Yes
  • Thinking: Yes
  • MCP: Yes
  • Coding: Yes
  • Structured output: Yes

Technical specifications

Provider
Meta
License
Proprietary
Context window
1M
Parameters
Undisclosed
Architecture
Transformer (proprietary)
Release
2026-09
Modalities
Text, Image, Video, Audio
Vision
Yes
Audio
Yes
Tool calling
Yes
Thinking
Yes
MCP
Yes
Open weights
No
API
Yes
Pricing (input)
~$1.25 per 1M tokens (Meta Model API / Muse Code standard)
Pricing (output)
~$4.25 per 1M tokens

Supported modalities

Text · Image · Video · Audio

Context window

1M (1,000,000 tokens)

Pricing

Input: ~$1.25 per 1M tokens (Meta Model API / Muse Code standard)
Output: ~$4.25 per 1M tokens

Contributor tier discounts may apply when sharing data for training. Verify developer.meta.com / Muse Code pricing.

Availability

API: Yes
Chat UI: Yes
Open weights: No
Regions: Meta Model API with expanded global access

Muse Spark 1.3 (2026-09-02) in Muse Code and Meta Model API. Previously available reasoning modes ship now; max reasoning is withheld pending additional safety testing. Open weights not released.

Use cases

  • Long-horizon agentic coding with Muse Code
  • Personal agentic orchestration across apps
  • Computer-use workflows
  • Multimodal perception + action loops

Strengths

  • 1.3 long-horizon agentic/coding upgrade vs 1.2 (fewer turns/tokens per Meta)
  • 1M context with context management
  • Multimodal + computer-use in one closed stack

Limitations

  • Closed weights (unlike Llama)
  • Newer ecosystem than OpenAI/Anthropic
  • Muse Code still in beta

Related guides

Related benchmarks

Related research

Related tools

Related rankings

Related comparisons

Companies

Explore more models

All models →