Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
OpenRouter logo

Model details

Muse Spark 1.2

Muse Spark 1.2 sits inside Meta's Muse family as a proprietary, reasoning-tuned release aimed squarely at complex agentic tasks rather than casual chat. It continues a rapid iteration cadence that began in April with Muse Spark 1.0, improved meaningfully in Muse Spark 1.1, and now adds stronger knowledge-work and coding capabilities. Independent evaluators describe it as Meta's strongest agentic offering to date, with measurable jumps on agent-oriented benchmarks that simulate real multi-step work such as preparing analyses and presentations. The model is positioned as a serious contender among US labs, closing the gap to the current frontier on agentic reasoning while retaining the multimodal input flexibility and long context that knowledge-work workflows demand.

Beyond agentic use, Muse Spark 1.2 shows particular strength on coding tasks relative to its other published evaluations, with review notes highlighting clean first-attempt performance across Python, TypeScript, and structured-data generation tasks. BenchLM identifies coding as its strongest eligible benchmark category, and Artificial Analysis records solid gains on agentic coding and terminal-use evaluations as well as a large lift on the GDPval-AA v2 agentic Elo. The practical fit is a model that can read across long, mixed-format documents, reason through multi-step problems with configurable effort, and execute tool-driven workflows in agentic loops, making it best suited to product teams building assistants, analysts, and developer tooling that need sustained reasoning rather than one-shot answers.

OpenRoutermeta/muse-spark-1.2muse

Quick Info

Powered by
Provider
OpenRouter
Model key
meta/muse-spark-1.2
Release date
Aug 5, 2026
Last updated
Aug 5, 2026
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$1.25
Output token cost
$4.25

Limits

Output tokens
943,718 tokens
Context window
1,048,576 tokens

Transparent token rates

Compare Muse Spark 1.2 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Muse Spark 1.2

Meta

Coverage

A primary-source dissection of Muse Code and Muse Spark 1.2 confirms that Meta announced both on August 5, 2026, with Muse Code positioned as a terminal-based coding agent and 1.2 as the model trained to run inside it. The analysis draws on the official Meta AI Research blog and the evaluation methodology report, organ On measurement, the article walks through Meta's actual benchmark figures for 1.2 rather than relying on chart images, and reports a case where GPU kernel optimization required roughly 24 hours of model time, illustrating the practical cost ceiling for long-horizon agentic runs on the standard tier. It also restates th

OpenRouter

Official sourceBenchmark

Muse Spark 1.2 is now available on OpenRouter as a reasoning model from Meta designed for complex agentic tasks. It accepts text, images, video, audio, and PDF documents, returns text, and provides a 1M-token context window. The model supports multi-agent workflows, configurable reasoning effort, structured output, and Pricing on OpenRouter is set at $1.25 per 1M input tokens and $4.25 per 1M output tokens, with the model released on August 5, 2026. OpenRouter forwards requests directly to Meta's single provider, with uptime, latency, and throughput metrics tracked on the listing page. The page also includes an OpenAI-compatible quic

Opper

Coverage

CNBC reported on August 10, 2026 that Meta CEO Mark Zuckerberg announced in an Instagram video that the company would open the weights of its latest model, Muse Spark 1.2, making the parameters downloadable and usable by the public. Zuckerberg framed the move as a competitive response to both U.S. closed-model labs lik The same announcement introduced Muse Glimmer, a new family of open-source models designed to run on laptops, positioned alongside the open-weight Muse Spark 1.2 release. The article places the decision within investor scrutiny of Meta's Superintelligence Labs capital expenditure plans, which Meta has forecast at up to

OpenRouter

CoverageBenchmark

An independent review of Muse Spark 1.2 confirms the August 5, 2026 launch and breaks out model capabilities from the Muse Code terminal agent. The model passed Python, TypeScript, and repository-wide JSONL coding tasks on first attempt via the API, while the review explicitly notes that Muse Code agent execution and t Artificial Analysis ranks Muse Spark 1.2 at 12 of 185 models with an Intelligence Index of 57, providing an independent cross-check on Meta's claims. The reviewer cautions that three controlled tasks do not establish agent reliability or tool-call quality, and recommends the Contributor tier's $0.10/$0.20 per million t

Meta

Coverage

Meta released Muse Spark 1.2 on August 5, 2026 alongside the beta of Muse Code, a terminal-based coding agent that Meta co-trained with the model rather than leaving harness development to the ecosystem. The published Meta Model API list rate is $1.25 per million input tokens and $4.25 per million output tokens on the Beyond pricing and packaging, the launch guide notes that Meta published benchmark charts for Muse Spark 1.2 only as images rather than as raw numbers in body text, which limits independent verification of headline scores and is a meaningful caveat for any team evaluating the model. The contributor-tier discount is exp

OpenRouter

Coverage

The Hacker News thread for Meta's Muse Code and Muse Spark 1.2 launch highlights a two-tier pricing structure on the Meta developer platform. The standard tier is priced at $1.25/M input and $4.25/M output, while the Contributor tier offers a 10x discount on input ($0.10/M) and a 20x discount on output ($0.20/M) if dev Commenters compared the Contributor opt-in to OpenAI's data-sharing program for complimentary daily tokens, noting Meta's pricing is simpler and more transparent. Thread participants debated the trade-offs of data usage for cost savings, with several users indicating they had enabled the tier for non-sensitive projects

OpenRouter

CoverageRelease Notes

Meta Superintelligence Labs released Muse Spark 1.2 on August 5, 2026, alongside Muse Code, a beta terminal coding agent co-trained with the model, according to Vorp Labs' release review. Standard-tier API pricing is unchanged from Muse Spark 1.1 at $1.25 per million input tokens, $0.15 cached input, and $4.25 per mill For developers evaluating access via OpenRouter, the Vorp Labs review explicitly notes that although Meta links an OpenRouter page for the model, no muse-spark-1.2 route existed in the OpenRouter catalog at the August 5 check, meaning the standard and contributor API tiers are accessible only through Meta's own API end

OpenRouter

Coverage

CNBC reports that Meta rolled out Muse Code on August 5, 2026 as the company's first AI coding agent, positioning it against Anthropic's Claude and OpenAI's Codex assistants. The agent works alongside the Muse Spark 1.2 model and is available in preview via a pay-as-you-go API option with pricing matching the prior Mus The CNBC piece frames Muse Code as part of CEO Mark Zuckerberg's broader push to generate revenue from AI investments following a light Q2 revenue forecast and concerns about free cash flow, with data center and computing infrastructure spending continuing at scale. Wang declined to share user statistics for the underl

OpenRouter

CoverageRelease Notes

Seeking Alpha's August 5, 2026 news brief confirms that Meta Platforms announced Muse Code as its first AI coding agent, released alongside the Muse Spark 1.2 model and framed competitively against OpenAI and Anthropic. The report positions the launch as Meta's entry into the autonomous coding-agent category that Anthr The full Seeking Alpha article sits behind a registration/paywall, so the supplied excerpt yields only headline-level detail without API pricing, context window, benchmark, or OpenRouter routing specifics. It is useful only as a timestamp and corroborating launch reference rather than as a primary technical source for

Opper

CoverageBenchmark

Artificial Analysis published a benchmark-driven breakdown of Muse Spark 1.2 on August 5, 2026, giving the model an Intelligence Index score of 54 (xhigh) — up three points from Muse Spark 1.1 and eleven points from the April Muse Spark 1.0 release. The piece credits Meta Superintelligence Labs for early access and pla Agentic knowledge-work gains drive the update: GDPval-AA v2 Elo rises 260 points to 1631 (fifth among all models benchmarked and ahead of Claude Opus 4.8), Terminal-Bench v2.1 improves from 78% to 80%, and τ³-Banking edges from 25% to 27%. The AA-Omniscience abstention rate climbs from 18 to 22 as hallucination falls 1

Merge Gateway

Coverage

Meta debuted Muse Spark 1.2 on August 5, 2026 as a coding-centric update to its Muse Spark model line, released through Meta Superintelligence Labs (MSL). According to Yahoo Finance, Spark 1.2 ships alongside Muse Code, the first coding agent from MSL, which serves as the harness running the model and positions Meta to The Muse Spark launch forms part of Meta's broader AI rebuild after its Llama models underwhelmed developers last year, an effort driven by chief AI officer Alexandr Wang following Scale AI's $14.3 billion acquihire. The release sequence spans Muse Spark, Spark 1.1, Spark 1.2, the Muse Image image generation model, and

Videos about Muse Spark 1.2

More models around Muse Spark 1.2