meta/muse-spark-open

模型发布 OpenRouter 模型 原文语言:中文

OpenRouter 于 2026-10-16 收录该模型(收录时间不等于模型正式发布日期);上下文:1,048,576 tokens;输入价:$1.2500/百万 tokens;输出价:$4.2500/百万 tokens;定位(来源描述):Muse Spark Open is a coding-optimized open-weight model.

全文

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

Compare

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$0.150.36s176 tok/s

99.88%

$1.25$4.25$0.150.37s202 tok/s

99.73%

$1.25$4.25$0.150.22s263 tok/s

100.00%

$1.25$4.25$0.151.09s84 tok/s

99.84%

$1.25$4.25$0.150.57s134 tok/s

99.97%

ProviderInput / Output /M

Input $1.25 / Output $4.25

176 tok/s99.88% uptime

Input $1.25 / Output $4.25

202 tok/s99.73% uptime

Input $1.25 / Output $4.25

263 tok/s100.00% uptime

Input $1.25 / Output $4.25

84 tok/s99.84% uptime

Input $1.25 / Output $4.25

134 tok/s99.97% 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

$0.3372/M tokens

Output

$4.249/M tokens

Muse Spark Open — Price history

Price history

No effective price data for this selection

$0.207$4.2594.8%51.8%

$0.342$4.2582.6%2.9%

$0.481$4.2569.9%36.4%

$0.514$4.2566.9%8.1%

$0.542$4.2564.3%0.8%

ProviderInput / Output /M

Token share

Effective in $0.207 / Effective out $4.25

Token share

51.8%

Effective in $0.342 / Effective out $4.25

Token share

2.9%

Effective in $0.481 / Effective out $4.25

Token share

36.4%

Effective in $0.514 / Effective out $4.25

Token share

8.1%

Effective in $0.542 / Effective out $4.25

Token share

0.8%

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

263tok/s

P50, best across providers

Latency

0.22s

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)The model was reachable. Request routed to a provider.

100.00%

Availability (3d)The model returned inference from any provider. Errors and empty responses count against it.

99.86%

Availability over the last 3 days

Last 72 hours

Availability 99.86%

3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability

99.96%

Without Routing

97.49%

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)

SourceBenchmarkScore

Artificial AnalysisMuse Spark Open (Xhigh) Intelligence Index24.9

Artificial AnalysisMuse Spark Open (Xhigh) GPQA Diamond89.0%

Artificial AnalysisMuse Spark Open (Xhigh) HLE43.9%

Artificial AnalysisMuse Spark Open (Xhigh) AA-LCR78.3%

Artificial AnalysisMuse Spark Open (Xhigh) GDPval-AA45.1%

Artificial AnalysisMuse Spark Open (Xhigh) CritPt14.9%

Artificial AnalysisMuse Spark Open (Xhigh) SciCode58.4%

Artificial AnalysisMuse Spark Open (Xhigh) Terminal-Bench 4.08.1%

Artificial AnalysisMuse Spark Open (Xhigh) Terminal-Bench 2.176.4%

Artificial AnalysisMuse Spark Open (Xhigh) AA-Omniscience Accuracy42.0%

Artificial AnalysisMuse Spark Open (Xhigh) AA-Omniscience Non-Hallucination Rate72.0%

Design ArenaMuse Spark Open (xhigh) Models Arena 3D Elo1235

Design ArenaMuse Spark Open (xhigh) Models Arena Code Categories Elo1299

Design ArenaMuse Spark Open (xhigh) Models Arena Game Development Elo1259

Design ArenaMuse Spark Open (xhigh) Models Arena UI Component Elo1309

Design ArenaMuse Spark Open (xhigh) Models Arena Website Elo1305

Vals AIMuse Spark Open Code Migration v1.126.9%

Vals AIMuse Spark Open EMB v1.162.0%

Vals AIMuse Spark Open Finance Agent v2.160.7%

Vals AIMuse Spark Open Vals Finance Index53.0%

Vals AIMuse Spark Open HLAB-Vals-1.125.0%

Vals AIMuse Spark Open Vals Legal Index68.5%

Vals AIMuse Spark Open Legal Research Bench38.5%

Vals AIMuse Spark Open SAGE85.8%

Vals AIMuse Spark Open Tax Agent Bench26.3%

Vals AIMuse Spark Open Terminal-Bench 4.06.6%

Vals AIMuse Spark Open Vals Index48.6%

Vals AIMuse Spark Open Vibe Code Bench v1.176.2%

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.

1.

pi

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

42.2Btokens

2.

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.

4.14Btokens

3.

Portkey AI

Control panel for AI apps

2.28Btokens

4.

Cursor

new

2.17Btokens

5.

atrium-flywheel-tier1

new

1.95Btokens

Rest of OpenRouter

4.65B

Top 5 Apps

27M

Activity

Token volume and request traffic to this model over time.

API

Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.

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

Frequently asked questions

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.

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.

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

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

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

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.

Muse Spark Open was released on October 15, 2026.

正文以来源语言显示。

参考来源

← 返回每日简报