pragma.vision Your technology observatory

Verification register AI & Agents

Readiness verdict

Liquid AI LFM2.5

A dated reading of what is claimed, reported, and independently verified in the current evidence.

As of
2026-06-28
Revision
1
Method
v1.0.0

Current reading

The readiness gap, in one scan

AI-assisted assembly · derived results

Claimed
80

Public ambition and stated capability

Reported
80

Observed practitioner reporting

Verified
74

Independently supported evidence

Gap
+6

Claimed minus verified

Evidence strength Strong

Decision

What the current evidence supports

Human editorial judgment · 2026-06-28

Low-friction candidate

Why
Open weights, broad runtime support, strong small-model instruction-following scores, and sub-1GB footprints make it cheap to trial; small knowledge capacity is mitigable with RAG and the main residual risk is unverified vendor benchmarks.
Next
Pull the open weights from Hugging Face and benchmark LFM2.5-1.2B-Instruct (and 8B-A1B for richer tasks) on target NPU/CPU hardware with retrieval for any knowledge-heavy path; for on-device reasoning/tool-use trial the 1.2B-Thinking variant; confirm the license matches intended commercial use.

Constraints

Blockers

No named blocker is present in the current public projection.

Evidence summary

Derived counts

AI-assisted assembly

Total
7
Tier 1
0
Tier 2
4
Tier 3
3
Supports
5
Contradicts
1
Context
1
Latest observed
2026-05-28

Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.

Publication record

Revisions

Initial public reading

This is the initial public reading. No earlier readiness change is recorded.

Your opinion

Tell us anything.

What works, what doesn't, what's missing — especially about our watches, lenses, and the register itself. Anonymous is fine; leave an email if you'd like a reply.