EN
Back to the archive

The archive · Space, Robots, Defence · Product decision · 2023–2026

MiRESSO bets fusion's bottleneck is beryllium: ¥4.23B Series A, QST-spinout pilot plant

QST spinout founded May 2023 bets low-temperature refining can make fusion-grade beryllium cheap; closed ¥4.23B Series A, pilot plant targeting 2027 production.

MiRESSO (ミレッソ)

The betThat fusion's bottleneck is beryllium, not reactors: a low-temperature refining process can make it cheap enough to anchor a supply chain before reactors are built.Building

What the business is

A Misawa, Aomori-based startup producing and selling beryllium and beryllium compounds for fusion reactors, plus licensing its low-temperature refining platform for other mineral refining and recycling.

Starting capitalSBIR Phase 3 grant of ¥2B from MEXT; Series A raised ¥1.83B (Aug 2025 first close) then ~¥2.4B (Feb 2026 final close) for ¥4.23B total; cumulative funding about ¥6.68B including grants as of February 2026.

How it started

Founded in May 2023 by Masaru Nakamichi, a researcher at QST's Rokkasho Fusion Energy Institute, the company was certified by QST and in October 2023 won MEXT's SBIR Phase 3 fusion program (¥2B). In October 2024 it signed a comprehensive partnership with Pacific Metals to build its BETA pilot plant at Pacific Metals' Hachinohe works.

What happened

In August 2025 MiRESSO closed an ¥1.83B Series A first close led by Spiral Capital with Pacific Metals and Genesia Ventures. In February 2026 it closed the final ~¥2.4B tranche with SBI Investment as co-lead and nine new investors including Mitsui & Co., NTT Docomo Ventures, Mitsubishi UFJ Capital, MPower Partners and Keio Innovation Initiative, completing a ¥4.23B Series A and bringing cumulative funding to about ¥6.68B.

How it ended up

Still private and building: with Series A money it is constructing BETA, a ~300m² beryllium pilot plant in Hachinohe, Aomori, targeting production start in fiscal 2027; Mitsui separately invested in March 2026 to help build the supply chain from ore procurement to product sales.

Background

MiRESSO was founded in May 2023 by Masaru Nakamichi, a researcher at QST's Rokkasho Fusion Energy Institute, with technology spun out of the national lab. Its bet is contrarian inside the fusion boom: rather than build a reactor, sell the neutron-multiplier material beryllium, without which tokamak blankets and plasma control cannot work — and refine it with a proprietary low-temperature process instead of today's costly methods.

The strategy moved from grants to plant fast. In October 2023 it won MEXT's SBIR Phase 3 fusion program worth ¥2B, and in October 2024 signed a comprehensive partnership with Pacific Metals to co-locate its pilot plant in Hachinohe. In August 2025 Spiral Capital led an ¥1.83B Series A first close; in February 2026 the round closed at ¥4.23B with SBI Investment co-leading and nine new investors — Mitsui, NTT Docomo Ventures, Mitsubishi UFJ Capital, MPower Partners and others — bringing cumulative funding to roughly ¥6.68B.

The money is going into BETA, a ~300m² beryllium pilot plant inside Pacific Metals' works, with production targeted in fiscal 2027. Mitsui's separate March 2026 investment underlines the supply-chain thesis: from ore procurement to refining to sales, MiRESSO wants to own the full chain for a material whose suppliers are currently few and concentrated overseas.

The bet is unproven — beryllium demand depends on fusion actually commercializing, and pilot-scale output in 2027 is a hard milestone. But the low-temperature process is also applicable to other rare metals and recycling, giving the platform business an option beyond fusion.

What has to be true

  • MiRESSO deliberately skipped the reactor race and bet on the material every reactor needs, positioning itself as fusion's supply chain rather than another device developer.
  • Spinning out of QST gave it laboratory-proven refining technology, de-risking the pitch to investors before any plant existed.
  • It compressed the path from grant to a first-of-a-kind pilot plant, letting Series A investors fund construction rather than just R&D.
  • Supply-chain and trading investors (Pacific Metals, Mitsui) came in early, validating both the technology and the go-to-market channel.
  • The refining platform extends beyond beryllium to other rare minerals and recycling, so the bet survives even if fusion slips.

What can be applied

In a hot technology race, the bottleneck material can beat the headline product: MiRESSO let reactor developers fight while it cornered the input everyone will need.

Aftermath

As of 1 September 2026, MiRESSO is in the build phase: Series A is closed at ¥4.23B, cumulative funding stands at roughly ¥6.68B including grants, and construction of the BETA beryllium pilot plant in Hachinohe is underway with production targeted in fiscal 2027. The company is also developing its mineral refining and recycling platform and actively hiring plant, R&D and business development staff.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe what you are building, and see who is betting on the same thing.

Free account · 3 free questions · no card

Related cases