Meta's Muse Spark 1.3 Claims #3 Global Ranking, Matches OpenAI's GPT-5.6-Sol on Coding Benchmarks
Meta Superintelligence Labs shipped Muse Spark 1.3, which the company claims ranks #3 globally on the Artificial Analysis Intelligence Index and matches OpenAI's GPT-5.6-Sol on coding and agentic benchmarks. The model is available now via Muse Code and Meta's API, with open weights and a follow-up model promised soon.
Meta's Muse Spark 1.3 Claims #3 Global Ranking, Matches OpenAI's GPT-5.6-Sol
Meta Superintelligence Labs released Muse Spark 1.3 on September 2, 2026, with CEO Mark Zuckerberg claiming the model delivers "frontier performance almost too cheap to meter" and represents "the biggest jump we've made so far on coding and agentic work."
According to the Artificial Analysis Intelligence Index (AAII), Muse Spark 1.3 now ranks #3 among all models globally — a claim that, if verified, would mark Meta Superintelligence Labs' first appearance among the top frontier labs alongside OpenAI and Anthropic. The model is reportedly matching OpenAI's GPT-5.6-Sol on comparative benchmarks, though Meta has not published detailed benchmark tables alongside the announcement.
What's confirmed
Muse Spark 1.3 is live now through Muse Code and Meta's API. Zuckerberg's announcement, posted to X, confirms the release date and positions the model as a major step up from prior Muse Spark versions, specifically on coding and agentic tasks. Meta says open-weight releases of Muse Spark are coming soon, alongside a next model teased only with a watermelon emoji — no name, specs, or timeline disclosed.
On pricing, Meta is reportedly offering a discount of more than 90% for developers who opt in to having their usage data used for training. Exact per-token pricing for either the standard or discounted tier has not been disclosed. Context window size, parameter count, and training data cutoff were also not specified in the announcement.
What's unverified
The #3 global ranking and the claim of matching GPT-5.6-Sol come from third-party aggregation (AAII) and social media commentary rather than a formal technical report or benchmark suite published by Meta. No independent verification of these rankings was available at time of writing. Readers should treat "frontier performance" and "biggest jump" as company claims until peer-reviewed or third-party benchmark runs confirm the standing.
What this means
If the AAII ranking holds up under scrutiny, this is a significant reversal for Meta's AI division, which has trailed OpenAI, Anthropic, and Google DeepMind on most public leaderboards for the past two years following the underwhelming reception of earlier Llama-generation models. A top-3 global ranking — even from a single index — would be the strongest signal yet that Meta Superintelligence Labs' reorganization and talent acquisition push is translating into competitive model quality.
The training-discount pricing structure is also notable: offering a 90%+ price cut in exchange for training-data rights suggests Meta is prioritizing data acquisition at scale over near-term revenue from this release, a strategy that mirrors early-stage pricing moves by other labs racing to build proprietary fine-tuning corpora from real-world usage.
The promise of open weights, if delivered, would be more consequential than the closed release itself — it would give Meta a genuine differentiator against OpenAI and Anthropic, both of which keep frontier-tier models closed. Until Meta publishes actual benchmark tables, model specs, and confirms the open-weight timeline, the #3 ranking should be read as an early, single-source signal rather than an established fact.
Related Articles
OpenAI Launches GPT-6 Astra, Says the Model May Already Qualify as AGI
OpenAI has released GPT-6 Astra, its most capable model yet, with benchmark scores the company says surpass GPT-5.6 Sol and Anthropic's Fable 5 models. President Greg Brockman called it a step into the 'AGI era,' though OpenAI acknowledges there's no agreed-upon threshold for that term.
Meta Releases Muse Spark 1.3 Contributor, a Low-Cost Multimodal Reasoning Model With 1M Context Window
Meta has released Muse Spark 1.3 Contributor, described as the cost-efficient contributor tier of its multimodal reasoning model line. The model offers a 1 million token context window at $0.10 per 1M input tokens and $0.20 per 1M output tokens, targeting experimentation and early-stage agentic workflows.
OpenAI Releases GPT-6 Astra, First Model to Cross 'Critical' Cybersecurity Threshold
OpenAI has begun rolling out GPT-6 Astra, the first model to reach the company's internal 'Critical' cybersecurity threshold. Access is being phased, with companies in OpenAI's Daybreak cybersecurity program getting priority following added safeguards after a prior model containment breach.
OpenAI Launches GPT-6 Astra, Matches Claude Fable Pricing at $10/$50 per Million Tokens
OpenAI has begun rolling out GPT-6 Astra, priced at $10/million input and $50/million output tokens to match Claude Fable. The model claims a 99.9% score on ARC-AGI 3 using a custom harness and leads on security and long-context benchmarks, though it trails Fable on general intelligence rankings.
Comments
Loading...