How much does Muse Spark Open cost?

Muse Spark Open costs $1.25/M input tokens and $4.25/M output tokens, with separate rates for Cache Read at $0.15/M tokens.

What is the context length of Muse Spark Open?

Muse Spark Open has a 1,048,576 token context window. It supports up to 262,144 completion tokens.

Does Muse Spark Open support tool calling and structured outputs?

Yes. Muse Spark Open accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.

What inputs and outputs does Muse Spark Open support?

Muse Spark Open accepts text, images, video, files such as PDFs, and audio as input and returns text.

Which providers serve Muse Spark Open?

Muse Spark Open is served by 5 providers on OpenRouter: Modal, Together, Nebius Token Factory, Fireworks, and Baseten. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

When was Muse Spark Open released?

Muse Spark Open was released on October 15, 2026.

Meta: Muse Spark Open

meta/muse-spark-open

Model weights

Compare

Muse Spark Open is a coding-optimized open-weight model. Trained specifically for agentic coding workflows, it delivers higher first-attempt accuracy, reliable tool calling, and deep codebase understanding across the repos developers actually work in.

Modalities

In / out price

$1.25 / $4.25per 1M

Context

1.0M

Released

Oct 15, 2026

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).

Latency / throughput


$1.25	$4.25	$0.15	0.36s	255 tok/s	100.00%
$1.25	$4.25	$0.15	0.91s	146 tok/s	99.89%
$1.25	$4.25	$0.15	0.94s	134 tok/s	99.58%
$1.25	$4.25	$0.15	2.21s	99 tok/s	99.64%
$1.25	$4.25	$0.15	0.66s	64 tok/s	99.86%

Input $1.25Output $4.25
255 tok/s100.00% uptime
Input $1.25Output $4.25
146 tok/s99.89% uptime
Input $1.25Output $4.25
134 tok/s99.58% uptime
Input $1.25Output $4.25
99 tok/s99.64% uptime
Input $1.25Output $4.25
64 tok/s99.86% uptime

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Weighted average

Input

$0.3268/M tokens

Output

$4.25/M tokens

Muse Spark Open — Price history

Price history

No effective price data for this selection

$0.286	$4.25	87.7%	29.6%
$0.320	$4.25	84.6%	18.3%
$0.329	$4.25	83.8%	37.7%
$0.383	$4.25	78.9%	9.5%
$0.485	$4.25	69.6%	4.9%

Effective in $0.286Effective out $4.25
Token share
29.6%
Effective in $0.320Effective out $4.25
Token share
18.3%
Effective in $0.329Effective out $4.25
Token share
37.7%
Effective in $0.383Effective out $4.25
Token share
9.5%
Effective in $0.485Effective out $4.25
Token share
4.9%

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Throughput

255tok/s

P50, best across providers

Latency

0.36s

P50, best provider

Reasoning effortLatency / throughput

AutoExacto benchmarks

GPQA DiamondTAU-BenchModal91.2%75.3%auto-routing90.1%75.8%Nebius Token Factory89.7%75.6%Fireworks90.0%74.9%Together89.7%73.3%Fireworks90.5%72.3%

+2 more providers

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Uptime (3d)

100.00%

Availability (3d)

99.34%

Availability over the last 3 days

Last 72 hours

Availability 99.34%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.91%

Without Routing

94.71%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.

Benchmarks

Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.

Benchmark score summary for Meta: Muse Spark Open (Artificial Analysis, Design Arena and Vals AI)
Source	Benchmark	Score
Artificial Analysis	Muse Spark Open (Xhigh) Intelligence Index	24.9
Artificial Analysis	Muse Spark Open (Xhigh) GPQA Diamond	89.0%
Artificial Analysis	Muse Spark Open (Xhigh) HLE	43.9%
Artificial Analysis	Muse Spark Open (Xhigh) AA-LCR	78.3%
Artificial Analysis	Muse Spark Open (Xhigh) GDPval-AA	45.1%
Artificial Analysis	Muse Spark Open (Xhigh) CritPt	14.9%
Artificial Analysis	Muse Spark Open (Xhigh) SciCode	58.4%
Artificial Analysis	Muse Spark Open (Xhigh) Terminal-Bench 4.0	8.1%
Artificial Analysis	Muse Spark Open (Xhigh) Terminal-Bench 2.1	76.4%
Artificial Analysis	Muse Spark Open (Xhigh) AA-Omniscience Accuracy	42.0%
Artificial Analysis	Muse Spark Open (Xhigh) AA-Omniscience Non-Hallucination Rate	72.0%
Design Arena	Muse Spark Open (xhigh) Models Arena 3D Elo	1248
Design Arena	Muse Spark Open (xhigh) Models Arena Code Categories Elo	1301
Design Arena	Muse Spark Open (xhigh) Models Arena Game Development Elo	1264
Design Arena	Muse Spark Open (xhigh) Models Arena UI Component Elo	1313
Design Arena	Muse Spark Open (xhigh) Models Arena Website Elo	1306
Vals AI	Muse Spark Open Code Migration v1.1	26.9%
Vals AI	Muse Spark Open EMB v1.1	62.0%
Vals AI	Muse Spark Open Finance Agent v2.1	60.7%
Vals AI	Muse Spark Open Vals Finance Index	53.0%
Vals AI	Muse Spark Open HLAB-Vals-1.1	25.0%
Vals AI	Muse Spark Open Vals Legal Index	68.5%
Vals AI	Muse Spark Open Legal Research Bench	38.5%
Vals AI	Muse Spark Open SAGE	85.8%
Vals AI	Muse Spark Open Tax Agent Bench	26.3%
Vals AI	Muse Spark Open Terminal-Bench 4.0	6.6%
Vals AI	Muse Spark Open Vals Index	48.6%
Vals AI	Muse Spark Open Vibe Code Bench v1.1	76.2%

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

There are many coding agents, but this one is yours.

42.2Btokens

Claude Code

Claude Code is Anthropic's agentic coding tool that reads your entire codebase, plans and executes changes across files, runs tests, and iterates on failures, all from natural language prompts.

4.16Btokens

Portkey AI

Control panel for AI apps

2.28Btokens

Cursor

new

2.19Btokens

atrium-pi-harness

new

1.95Btokens

Rest of OpenRouter

8.03B

Top 5 Apps

27.3M

Not currently available through EU in-region routing. In-region routing docs

More models from Meta

Muse Ember

Muse Ember is the smallest and most efficient model from Meta that’s designed for complex reasoning execution and the work agents repeat all day: tool calls, coding agent subtasks, routing, extraction, and long document parsing.

Muse Ember accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context window.

Text1.0M context$0.10 / $0.50

Muse Spark Open

Text1.0M context$1.25 / $4.25

Muse Wonder

Muse Wonder is a frontier reasoning model from Meta, designed for complex personal and professional agentic tasks. It accepts text, images, video, audio, and PDF documents and returns text, with a 1M-token context window.

The model handles long-horizon work: it plans, delegates to sub-agents, calls tools in parallel, and recovers from errors, so multi-step tasks run to completion without step-by-step direction. It can also operate browsers and desktop and mobile applications, and works across many files in existing repositories to make and verify changes. Muse Wonder carries important context across conversations and retrieves the details relevant to the task at hand. And it's trained for Active Focus, so it works toward your goal even when a task gets challenging.

In Meta's testing, it performs well on complex engineering tasks, extended agentic sessions, and multi-file refactors to tackle developers' most demanding work.

Text1.0M context$4 / $20

Muse Voice Transcribe 1.0

Muse Voice Transcribe 1.0 is a synchronous speech-to-text model from Meta. It is suited for push-to-talk, endpointing, and speaker-aware transcription, with keyword biasing for domain terms and language biasing through language-name hints. It accepts mono 16-bit PCM WAV audio at 16 kHz or 24 kHz for recordings up to 10 minutes. It does not provide word-level timestamps or confidence scores, and other audio formats must be converted to WAV before upload.

Transcription$0.00005/second

Muse Spark 1.3 Contributor

Muse Spark 1.3 Contributor is the cost-efficient contributor tier of Meta’s multimodal reasoning model for experimentation, learning, and early-stage agentic, multi-agent, and coding workflows. It is designed to track information across extended tasks, work through conflicting inputs, and request clarification or confirmation when needed. Prompts and outputs may be used to improve Meta’s products.

Text1.0M context$0.10 / $0.20

Muse Spark 1.3

Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It is designed to keep track of information across extended tasks, work through conflicting inputs, and request clarification or confirmation when needed, with an emphasis on concise execution.

Text1.0M context$1.25 / $4.25

Muse Image

Muse Image is an agentic image generation model from Meta that generates and edits images from text and reference images. Unlike single-pass image models, it reasons before it renders, breaking down multi-part prompts and refining its output within the chain of thought, and invokes web search for factual accuracy on knowledge-intensive prompts.

The model supports text-to-image generation, targeted image editing, multi-image composition, reference-image conditioning for style and subject consistency across a series, and precise text rendering within generated images. Iterative editing works by passing the previous output image back with a new instruction.

Image$0.01/image

Muse Spark 1.2 Contributor

Muse Spark 1.2 contributor tier is a reasoning model from Meta designed for developers who want to start building at an even lower cost. It’s meaningfully cheaper than Muse Spark 1.2. Your prompts and outputs may be used to improve Meta’s products, making it ideal for experimentation, learning, and early-stage projects without worry about spend.

It is a reasoning model from Meta, tailored for complex agentic tasks. It accepts text, images, video, and PDF documents, returns text, and offers a 1M-token context window. The model is built to support multi-agent workflows, whether as either a main agent that plans and delegates or as a subagent executing in parallel. It works across multiple coding harnesses and supports structured output, parallel function calling, and configurable reasoning effort. In Meta’s testing, it performs well on multi-file refactors, extended debugging sessions, whole-repository generation, and tasks that stretch well past a single prompt.

Text1.0M context$0.10 / $0.20

Muse Glimmer 30B

Muse Glimmer 30B is a dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware. It is suited for long-horizon agentic and coding workflows, with multi-step reasoning, reliable tool use, failure recovery, image understanding, and multilingual support across more than 100 languages.

Text131K context$0.30 / $1.20

Muse Glimmer 30B (batch)

Text131K context$0.24 / $0.96

Muse Spark 1.2

Muse Spark 1.2 is a reasoning model from Meta, designed for complex agentic tasks. It accepts text, images, video, and PDF documents, returns text, and offers a 1M-token context window.

The model is built to support multi-agent workflows, whether as either a main agent that plans and delegates or as a subagent executing in parallel. It works across multiple coding harnesses and supports structured output, parallel function calling, and configurable reasoning effort. In Meta’s testing, it performs well on multi-file refactors, extended debugging sessions, whole-repository generation, and tasks that stretch well past a single prompt.

Text1.0M context$1.25 / $4.25

Muse Spark 1.1

Muse Spark 1.1 is a multimodal reasoning model from Meta, built for agentic tasks. It accepts text, images, video, and PDF documents and returns text, with a 1M-token context window.

The model is designed to orchestrate multi-agent workflows, acting as either a main agent that plans and delegates or as a subagent, and generalizes zero-shot to new tools, MCP servers, and custom skills. It supports structured output, parallel function calling, built-in search with citations, and configurable reasoning effort. Meta reports strong performance on real-world coding across large codebases, computer-use workflows, and visual-to-code generation.

Text1.0M context$1.25 / $4.25

Meta: Muse Spark Open

Providers