Meta Releases Muse Spark 1.3, a Free Multimodal Reasoning Model with 1M-Token Context
Meta has released Muse Spark 1.3, a multimodal reasoning model with a 1M-token context window, listed as free on OpenRouter. The model targets long-running agentic, multi-agent, and coding workflows, though audio input support remains incomplete.
Meta: Muse Spark 1.3 — Quick Specs
Meta Releases Muse Spark 1.3
Meta has released Muse Spark 1.3, a multimodal reasoning model designed for long-running agentic, multi-agent, and coding workflows. The model is now listed on OpenRouter with a 1 million token context window and is currently available at no cost.
What Muse Spark 1.3 Does
According to Meta, Muse Spark 1.3 is built to maintain state across extended tasks, reconcile conflicting inputs, and pause to request clarification or confirmation from users when needed. The model emphasizes concise execution over verbose output, positioning it for use in autonomous and semi-autonomous agent pipelines rather than single-turn chat.
The 1M-token context window puts Muse Spark 1.3 in the same range as other long-context models designed to hold entire codebases, multi-document research sets, or extended multi-agent transcripts in a single session.
Known Limitations
Meta's own listing flags a specific gap: audio understanding in Muse Spark 1.3 is "currently not fully supported," and response quality on requests that include audio content may be degraded. This suggests the model's multimodal capabilities are more mature for text and likely image/document inputs than for audio, and teams building audio-dependent pipelines should treat this as a known constraint rather than a fully shipped feature.
Pricing and Availability
Muse Spark 1.3 is currently listed as free for both input and output tokens on OpenRouter, with no per-token cost disclosed beyond the zero-price listing. OpenRouter reports 100% uptime over the trailing three days and 99.69% availability over the last 24 hours across its routing infrastructure, though these figures reflect platform-level routing performance rather than model quality or accuracy.
No benchmark scores, parameter count, or training data cutoff date have been disclosed by Meta for this release. OpenRouter's listing shows a release date of September 2, 2026; readers should treat this date as sourced directly from the OpenRouter model card, as Meta has not published an independent announcement with corroborating details at the time of writing.
What This Means
A free, 1M-token multimodal reasoning model lowers the barrier for developers experimenting with long-running agent workflows, particularly multi-agent systems that need to retain context across many steps without expensive re-summarization. The lack of published benchmarks makes it difficult to assess how Muse Spark 1.3 compares to other reasoning models like GPT or Gemini variants on standard coding and reasoning suites — buyers evaluating it for production agentic pipelines should run their own evals rather than relying on Meta's framing. The explicit audio limitation is a useful signal: this is a model still being hardened for full multimodal input, and teams should scope their initial use to text, code, and likely image-based tasks until Meta confirms audio support is production-ready.
Related Articles
DeepSeek V4.1 Flash 1017 launches with 552B CED backbone, 1M context, $0.15/$0.60 per 1M tokens
DeepSeek has released V4.1 Flash 1017, an October 17, 2026 snapshot built on its new Causal Encoder-Decoder architecture. It uses a 552B-parameter sparse MoE backbone with a 1M-token context window, native image understanding, and pricing of $0.15 input / $0.60 output per 1M tokens on OpenRouter.
Moondream 3 Preview lists on OpenRouter: 9B MoE vision model, 2B active, 32K context, $0.30/$2.50
Moondream 3 Preview is a vision-language model with 9B total and 2B active parameters (mixture-of-experts), a 32K-token context window and an optional reasoning mode. OpenRouter lists it at $0.30 per 1M input tokens and $2.50 per 1M output tokens, served by Prime Intellect.
Moondream 3.1: 9B MoE vision model with 2B active parameters, 32K context, $0.30/$2.50 per 1M tokens
Moondream has released Moondream 3.1, a mixture-of-experts vision-language model with 9B total and 2B active parameters and a 32K-token context window. It is listed on OpenRouter at $0.30 per 1M input tokens and $2.50 per 1M output tokens, with an optional reasoning mode.
Unbiased releases Pareto 26.9: 262K-context composite model at $2.50/$7.50 per 1M tokens
Unbiased has released Pareto 26.9, a multimodal composite model with a 262K-token context window, priced at $2.50 per 1M input tokens and $7.50 per 1M output tokens. It is listed on OpenRouter with a Oct 13, 2026 release date. A cheaper 1.0M-context preview, Pareto 26.10, is already listed alongside it.
Comments
Loading...