Meta launches Muse Spark 1.1 coding model at $1.25/$4.25 per million tokens
Meta publicly released Muse Spark 1.1, a multimodal AI model designed for agentic coding workflows. The model is priced at $1.25 per million input tokens and $4.25 per million output tokens, positioning it slightly above Anthropic's Claude Haiku 4.5 and OpenAI's GPT-5.6 Luna.
Muse Spark 1.1 — Quick Specs
Meta launches Muse Spark 1.1 coding model at $1.25/$4.25 per million tokens
Meta publicly released Muse Spark 1.1 on Thursday, a multimodal AI model designed for agentic coding that competes directly with offerings from OpenAI and Anthropic. The model is priced at $1.25 per million input tokens and $4.25 per million output tokens.
Pricing and positioning
Meta's pricing positions Spark 1.1 slightly above Anthropic's Claude Haiku 4.5 and OpenAI's GPT-5.6 Luna, according to Reuters. The company claims the model excels at agentic performance, tool use, and computer use — capabilities increasingly demanded by enterprises for automation.
The first version of Muse Spark was announced in April 2025. Meta describes Spark 1.1 as capable of multi-step reasoning, managing digital workflows, and deploying features in enterprise systems. The company specifically targets use cases like handling large agentic workloads, bug fixes, and large-scale code migrations.
Technical capabilities
According to Meta's blog post, "Muse Spark 1.1 delivers exceptional performance in personal agentic tasks that require planning and orchestration across a range of external apps and services." The model is described as multimodal, though specific modalities beyond text and code have not been detailed.
Context window size, benchmark scores, and parameter count have not been disclosed.
Market context
Meta enters the agentic coding market behind its competitors. Anthropic and OpenAI have offered similar capabilities for an extended period. However, CEO Mark Zuckerberg broke a three-year absence from X (formerly Twitter) to announce the release, calling it "a strong agentic and coding model at a very low price." His last post was in July 2023.
Zuckerberg indicated "more to come soon," suggesting additional model releases are planned.
What this means
Meta's entry into agentic coding intensifies competition in a crowded field where pricing has become a key differentiator. While the company arrives late, its competitive pricing and focus on enterprise automation workflows — bug fixing, code migrations, and multi-app orchestration — addresses real enterprise pain points. The release coincides with a busy week of AI announcements, including GPT-5.6 from OpenAI and Grok updates from xAI, indicating accelerated release cycles across the industry. Meta's willingness to price aggressively suggests it's leveraging its infrastructure scale to compete on cost, a strategy that could pressure profit margins across the sector.
Related Articles
NVIDIA Nemotron 3.5 Lightning Arrives on Amazon SageMaker JumpStart, Targets High-Volume Agentic Workloads
NVIDIA's Nemotron 3.5 Lightning, a 30B-parameter hybrid Mixture-of-Experts model with only 3B active parameters, is now available for one-click deployment on Amazon SageMaker JumpStart. NVIDIA claims up to 4x higher throughput and 30% faster task completion for high-volume agentic workloads compared to larger frontier models.
Alibaba Releases Qwen3.8-27B-FP8, a 27B Dense Vision-Language Model with 1M-Token Context
Alibaba's Qwen team has released FP8-quantized weights for Qwen3.8-27B, a 27-billion-parameter dense vision-language model with native 262,144-token context extensible to 1 million tokens. The model claims gains over its Qwen3.6 and Qwen3.7 predecessors on coding, agentic, and multimodal benchmarks.
Google DeepMind Ships Gemini 3.7 Flash, Closing Gap With Claude 4.8 and GPT-5.5
Google DeepMind has released Gemini 3.7 Flash, a new entry in its fast-tier model line that reportedly closes a performance gap that opened up under Gemini 3.5 and 3.6 Flash against Anthropic's Claude 4.8+ and OpenAI's GPT-5.5+ series. Full pricing and benchmark details have not yet been disclosed.
Writer Launches Palmyra X6, an Open-Source-Based Model Aimed at Cutting Token Costs 50%
Writer released Palmyra X6, a post-trained variant of Z.ai's open-source GLM-5.2 model, alongside an upgraded agentic harness. The company claims the combination can cut customer token costs by as much as 50% for basic tasks.
Comments
Loading...