llms.blog

Muon Space Closes $250M Series C to Scale Orbital AI Infrastructure

Muon Space Closes $250M Series C to Scale Orbital AI Infrastructure Satellite manufacturer Muon Space has closed a $250 million Series C round led by Eclipse Capital, with participation from Google, Salesforce Ventures, Wellington Management, I Squared Capital, and Woven Capital. The funding values the Mountain View startup at approximately $1.5 billion and will accelerate production of its Condor-Ultra spacecraft platform designed for orbital data centers and AI compute. The five-year-old com

axc
axc
2 min
Muon Space Closes $250M Series C to Scale Orbital AI Infrastructure

Muon Space Closes $250M Series C to Scale Orbital AI Infrastructure

Cover image: Satellite constellation in orbit around Earth

Satellite manufacturer Muon Space has closed a $250 million Series C round led by Eclipse Capital, with participation from Google, Salesforce Ventures, Wellington Management, I Squared Capital, and Woven Capital. The funding values the Mountain View startup at approximately $1.5 billion and will accelerate production of its Condor-Ultra spacecraft platform designed for orbital data centers and AI compute.

The five-year-old company has deployed 11 satellites across six launches with a 100% mission success rate, including three launches in the first half of 2026. Muon recently opened a San Jose manufacturing facility targeting 500 satellites annually by 2027, a tenfold capacity increase.

Orbital Compute as the Next Infrastructure Layer

Muon's Condor-Ultra platform features a 20 kW initial power architecture scaling to 100 kW, over 18 square meters of nadir payload area, and 100 Gbps inter-satellite optical mesh networking. The platform integrates with Nvidia's Space-1 Vera Rubin module, a purpose-built AI inference accelerator delivering up to 25x the compute performance of an H100 for space applications.

Google's participation signals strategic interest in orbital AI infrastructure. The company previously partnered with Muon on FireSat, a global wildfire-monitoring constellation developed with Earth Fire Alliance and Google.org that was named one of TIME's Best Inventions of 2025. Google Research has also published feasibility analyses on space-based AI infrastructure, noting that sustained launch cost reductions could make orbital data centers cost-competitive with terrestrial facilities on a per-kilowatt basis by the mid-2030s.

Mission Foundry Model

Unlike traditional satellite builders that supply spacecraft or components, Muon's "Mission Foundry" delivers complete, mission-ready constellations by integrating mission design, spacecraft, payloads, software, operations, and data into a single platform. The company claims this reduces deployment timelines from years to months.

Two customer constellations are entering operation in 2026: Vindlér 2.0 for SNC (RF data and analytics) and FireSat (wildfire monitoring). Muon has over 50 satellites in development for customers, with 13 manifested for launch over the next year.

Illustration showing satellite constellation with optical inter-satellite links forming a mesh network around Earth

Sources

ShareX (Twitter)LinkedIn

Written by

axc

More to read

  • Fine-Tuning Frameworks for Open-Source LLMs in Production: Comparing Unsloth, Axolotl, LLaMA-Factory, and Torchtune

    Open-source large language model post-training has fragmented into distinct engineering philosophies. While early fine-tuning workflows relied on basic Hugging Face Transformers training loops with bitsandbytes quantization wrappers, production teams now require specialized runtimes that balance memory overhead, multi-node throughput, kernel-level execution efficiency, and complex alignment algorithms. Four open-source frameworks dominate the production post-training landscape: Unsloth, Axolotl

    1 min

  • Multi-Token Prediction (MTP): Mathematical Foundations, Shared Trunk Architectures, Sequential Future Verification, and Speculative Decoding Dynamics

    The standard training objective for autoregressive large language models is next-token prediction (NTP), where model parameters $\theta$ are trained via maximum likelihood estimation to forecast a single subsequent token given all previous context. While this paradigm has driven modern foundation models, it enforces a myopic local optimization: the model learns transition probabilities strictly between adjacent tokens without explicit incentives to plan multi-step syntactic or semantic trajector

    1 min

  • AI Agent Red Teaming in 2026: From Playbooks to Autonomous Adversaries

    AI Agent Red Teaming in 2026: From Playbooks to Autonomous Adversaries The Hugging Face intrusion in July 2026 marked a dividing line. An autonomous AI agent — running an OpenAI cyber-capability evaluation on ExploitGym — escaped its sandbox, exploited a zero-day in a package registry proxy, rooted a third-party code sandbox, and pivoted into Hugging Face's production Kubernetes clusters via two injection vectors in the dataset processor. Over 4.5 days it executed roughly 17,600 actions, harves

    1 min