New FRAND Guidelines of the 7th Civil Chamber of the Munich I Regional Court

The Munich patent chamber that decides the bulk of German SEP cases has set out, over 67 pages, how it assesses FRAND. Anyone pleading in Munich now knows which figures the chamber expects and which lines of defense it will not accept.

Background

The 7th Civil Chamber of the Munich I Regional Court, chaired by Dr. Oliver Schön and sitting with Dr. Florian Schweyer and Katalin Tözsér, issued new FRAND Guidelines dated 13 August 2026. They consolidate the chamber’s approach from its recent SEP decisions and add points so far discussed only in unpublished proceedings. The presiding judge explains the document by reference to a “clear need for a systematic and clearly structured presentation in practice” and to the aim of deciding cases within one year of filing.

The Guidelines supersede the orientation previously provided by the joint February 2020 notice of the 7th and 21st Civil Chambers on the handling of the competition-law compulsory-license defense under Huawei v ZTE, which was still closely tailored to the implementer working through the Huawei steps. The 21st Civil Chamber has been informed of the new guidance; given the level of detail, joint guidance was evidently not feasible. The Guidelines are not binding law and expressly reflect the position of the 7th Civil Chamber alone. Their practical weight follows from the number of cases pending there — some 40 to 50 new SEP cases a year — and from the announcement that the chamber will more often decide on the appropriateness of the claimant’s offer itself.

Procedure: “FRAND FIRST” Instead of a Safe Harbor

The most important procedural innovation is the option of a “FRAND FIRST” hearing: either party may request an early hearing devoted exclusively to FRAND issues, with submissions confined to a short brief of no more than 25 pages plus annexes. The chamber thereby continues the separation of infringement and FRAND observed since 2022 without abandoning the Munich procedural framework.

A genuine safe harbor is expressly rejected: a mechanism shielding an implementer from an injunction where it submits to binding arbitration would, in the chamber’s view, be incompatible with the patent holder’s interests and open to abuse. Providing security alone likewise does not prevent an injunction. The chamber sees no fundamental change in the establishment of the UPC Patent Mediation and Arbitration Centre in June 2026: mediation has been available at the Munich I Regional Court for more than 20 years and arbitration before bodies such as the ICC has long been available; all that has been added is a further option.

Willingness to License: External and Internal

The chamber distinguishes external from internal willingness. External willingness is assessed in a formalized way: the implementer must pay the undisputed portion of the royalty — measured by its own final offer — and, in defined circumstances, provide additional security. Under the chamber’s decisions, that includes, where the offers are far apart, security in the amount of the royalty falling due for one year of the license period under the claimant’s offer; and where the implementer is pursuing rate-setting proceedings abroad and that court has already proposed an amount, security in that amount. Only once external willingness is established does the chamber examine whether the SEP holder’s offer falls within the FRAND range; an implementer that nevertheless declines such an offer lacks internal willingness.

Determining the Range: Comparable Licenses First

The primary tool for determining the range is comparable licenses of the same licensor covering materially the same portfolio. The top-down approach serves as a cross-check in principle, and as the basis for an initial rate determination only where there is no established licensing practice or no suitable comparable agreement.

On comparability, the Guidelines provide in essence:

Agreements more than five years old are generally out of consideration.
Single-standard licenses are to be compared with single-standard licenses; multi-standard agreements generally cannot be unpacked, and cross-licenses are generally unsuitable unless the consideration can be valued independently.
Volume, duration and — exceptionally — geographic adjustments are possible; for exceptionally large volumes, discounts of up to around 30% are contemplated.
From the agreements submitted the chamber derives a median; the range extends 50% above and below it. Where a directly comparable agreement exists, the holder may increase the rate by no more than 15% — not cumulatively, but always measured against the original reference contract.

The top-down cross-check works with standardized device prices rather than actual selling prices (around US$170 for mobile phones) and with an aggregate royalty burden (around 8% for 5G). It presupposes a substantial share of the standard-relevant patents on the holder’s side; for mobile standards, around 1% may already suffice.

Court-appointed experts for rate determination are rejected: the appropriateness of the rate is a question of law and not amenable to expert evidence. Party experts, by contrast, may assist with calculations, the normalization of lump sums, options, releases, patent quality and top-down assumptions.

Specific Fields: Avanci 5G, Streaming, Past Use

The Guidelines then go on to discuss the chamber’s own decided cases in detail and to explain why it considered the terms at issue there appropriate. At least for the technical fields and market environments concerned (WiFi and HEVC licensing, mobile communications, automotive via the Avanci 5G pool, and “streaming as a service“) this yields very useful and in part highly concrete calculations; for the streaming environment, where no established licensing practice yet exists, the chamber offers an expressly non-binding illustrative calculation. It also addresses compensation for past use (“past release”), portfolio splitting, and a discount on the portfolios of Chinese companies (expressly presumed to be rebuttable, and likely open to challenge).

Anyone pleading in one of these fields will now find in the Guidelines a reliable indication of the orders of magnitude the chamber considers appropriate, and of the reasoning by which it arrives at them.

Practical Implications

SEP holders must prepare comparable licenses early and completely (licensor, portfolio, standard coverage, term, volumes) and keep increases over the reference contract within the limits set out in the Guidelines. The top-down cross-check with standardized device prices must always be run alongside.

Implementers must substantiate external willingness by payment: the undisputed portion measured by their own final offer, on a permanent basis, plus additional security where their own offer falls below the demand within the margins identified in the Guidelines. Relying solely on security, or on an offer to arbitrate, provides no protection against an injunction.

Both sides should bring in economic expertise early (but as party submissions, not in the expectation of a court-appointed expert).

Case planning: a “FRAND FIRST” hearing can save costs but forces an early, robust valuation; a party requesting one should already have its numbers ready within the page limit set by the Guidelines. After filing, it typically takes 9 to 12 months to reach an oral hearing.

Conclusion

As was to be expected, the chamber shifts the focus further still from conduct (the “behavioural approach”) to valuation: what decides a case in Munich is no longer only whether the implementer formally worked through the Huawei steps, but whether the holder’s offer falls within a transparently derived range. That creates predictability while raising the pleading burden on both sides. Whether the specific figures survive review by the Munich Higher Regional Court and the Federal Court of Justice remains to be seen.