SHELFCRITTER AI Digest

COMPANY

Reflection

companytopic-note

Overview

Reflection is one of eight companies designated by the Pentagon for classified-network AI deployment at Impact Level 6 (Secret) and Impact Level 7 (Top Secret) as of May 2026. The company enters the public record as a DOD partner for IL6/IL7 classified-network AI infrastructure deployments.

Timeline

  • 2026-05-02-AI-Digest — Pentagon signs classified-network AI deployment agreements with eight companies: OpenAI, Google, Microsoft, Amazon, NVIDIA, SpaceX, Oracle, and Reflection. Deployments cover IL6/IL7 classified DoD networks. Anthropic is conspicuously absent from the list.

  • 2026-07-01-AI-Digest — Open-weights lab Reflection (founded 2024 by ex-DeepMind researchers Misha Laskin and Ioannis Antonoglou; ~$25B raise reportedly in-progress) signs a $6.3B compute-lease deal with SpaceX — $150M/month starting July 1, 2026 through 2029 for access to NVIDIA GB300 systems at the Colossus 2 data centre near Memphis. 90-day mutual exit clause after month 3 makes the take-or-pay portion much smaller than the headline number. NVIDIA sits on both sides — $800M investor in Reflection and GB300 supplier for Colossus 2. First clear public case of a labs-tier open-weights customer landing a frontier-tier GB300 lease via a hyperscaler-adjacent counterparty, and the first public instance of GB300 supply pacing an open-weights lab (previously the constraint was only visible at closed frontier labs).

  • 2026-10-06-AI-Digest — Reflection unveiled Beam, a 501B-parameter sparse Mixture-of-Experts model (23B active per token) aimed at coding, reasoning, and agentic workloads. Weights ship under Apache 2.0 later in October, with a technical report plus the full fine-tuning and eval stack. Load-bearing softener carried from the digest: Reflection’s “3–4× less inference compute than GLM 5.2” claim is their own approximate comparison, excluding prompt processing, attention operations that scale with context, and serving overhead — not an independent benchmark. Capital-shape clarification worth pinning: Reflection took an NVIDIA-led $2B round in Oct 2025 at $8B valuation (Sequoia, Lightspeed, DST, 1789, Eric Schmidt as co-investors), separate from the SpaceX Colossus 2 GPU-compute contract; a ~$25B pre-money round reportedly closed in April 2026. Reframe worth carrying: Beam is a Western lab shipping Apache-2.0 weights at Chinese-model compute efficiency, not a cadence intensification of 500B+ open-weight releases — the Jul–Sep cohort (Kimi K3 2.8T, Qwen 3.8 2.4T, DeepSeek V4 Flash 284B) already carries the cadence thread, and Beam at 501B/23B active sits on the smaller end of it. Log against MOC - Open Source Models and MOC - AI Infrastructure.

  • 2026-10-11-AI-Digest — Per Bloomberg surfacing an FT report, NVIDIA is in early-stage talks either to boost its equity stake in Reflection or to supply the lab with compute capacity. NVIDIA already holds ~$800M from the Oct 2025 $8B-valuation round; Reflection is simultaneously in talks for a $2.5B raise at a $25B valuation per prior WSJ reporting, and has a $1B+ Nebius GB300 compute pact plus a multi-billion SpaceX compute deal (July 2026). Load-bearing softener: the headline framing is equity-stake-up vs compute-supply talks, not acquisition — the public-record reads as the former. No specific dollar figure for the Oct-10 tranche has been disclosed. Reframe worth carrying: the compute-supply arm is non-dilutive for Reflection but makes NVIDIA's circular-financing pattern sharper — pattern pairs with Jan 2026 $2B CoreWeave equity plus $6B+ pre-booked service commitment through 2032 and OpenAI‘s reported $100B commitment plus OpenAI‘s $22.4B CoreWeave contract. Log against MOC - Major Companies and MOC - AI Infrastructure.

  • 2026-10-13-AI-Digest — Per the FT (via Bloomberg and secondary coverage), NVIDIA is in early talks to deepen its position in Reflection — either by increasing its equity stake or by acquiring the lab outright, with acqui-hire structures reportedly on the table to dodge antitrust scrutiny. NVIDIA is already a backer with roughly $800M committed; Reflection’s March 2026 round priced it at $25B post-money. No deal terms are set, no price is public, and the “compute-supply alternative” framing circulating in some earlier summaries does not appear in the primary FT reporting. The talks surface five days after Reflection announced its first frontier open-weight model Beam. Load-bearing softener: structure is explicitly fluid — not an announced acquisition, and acqui-hire framing (if real) would price the deal very differently from a full-equity buyout. Reframe worth carrying: [[NVIDIA]] expanding its position in a US open-weights lab it already supplies with chips, not [[NVIDIA]] acquiring a model lab. Watch: whether terms leak into Reuters or Bloomberg direct in the next week, and whether any CMA / FTC pre-notice surfaces. Log against MOC - Major Companies and MOC - AI Infrastructure.

Key Developments

  1. Pentagon Classified-Network Deployment: Selected as one of eight strategic partners for IL6/IL7 classified DoD network AI deployments, indicating Pentagon confidence in the company’s security posture and AI capabilities for national-security applications.

  2. $6.3B SpaceX Colossus 2 Compute-Lease (July 1, 2026): Payment starts July 1, 2026 at $150M/month through 2029 for GB300 access at Colossus 2 — the campus originally built for xAI and folded into SpaceX after Musk’s absorption of xAI. Reflection is the second labs-tier customer to lease Colossus stack capacity (after Anthropic‘s May 2026 Colossus 1 lease) and the first open-weights lab to sit on GB300 supply at this scale. NVIDIA on both sides ($800M investor + GB300 supplier) is the structural signal — supply chain, capital chain, and customer chain converging on a single vehicle.

Updated 2026-07-01