liquid
LiquidAI: LFM2.5-Embedding-350M (free)
liquid/lfm-2.5-embedding-350m:free
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
liquid/lfm-2.5-embedding-350m:free is the OpenRouter free listing for Liquid AI's LFM2.5-Embedding-350M, a dense bi-encoder for short multilingual retrieval. The catalog maps it to the corresponding Hugging Face model, whose documentation reports 1,024-dimensional document vectors, a 512-token limit, cosine similarity, and support for 11 languages. Vendor-reported tables cover multilingual retrieval and query-embedding latency; OpenRouter says requests and embeddings may be retained. Validate the endpoint on your workload.[1][2]
Strengths
- Dense bi-encoder design: one 1,024-dimensional vector per document, scored with cosine similarity.[1][2]
- Multilingual and cross-lingual retrieval across 11 documented languages.[2]
- The vendor reports multilingual NanoBEIR and MKQA measurements plus low-latency query embedding in tested llama.cpp and enterprise GPU environments.[2]
Limitations
- The 512-token limit targets short passages; long documents need chunking or another model.[1][2]
- Direct use requires separate query: and document: prefixes; omission can silently reduce retrieval quality, and this pack does not show how the OpenRouter endpoint applies them.[1][2]
- Evidence is limited to OpenRouter metadata and Liquid AI documentation; no independent validation was supplied. OpenRouter says successful requests and embeddings may be retained for training and lists an expiration date.[1][2]
Where to use it
- Consider for multilingual e-commerce product search when product text fits the context limit.[2]
- Consider for FAQ or support knowledge-base retrieval in RAG pipelines after testing answer quality on your content.[2]
- A possible choice for semantic search across files, email, notes, or enterprise documents in supported languages, subject to retention and chunking checks.[1][2]
Sources
- [1] OpenRouter catalog · LiquidAI: LFM2.5-Embedding-350M (free)Sources checked: Nov 13, 2026
- [2] Model documentation · LiquidAI: LFM2.5-Embedding-350M (free)Sources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
LFM2.5-Embedding-350M is a text embedding model from Liquid AI. It produces 1,024-dimensional embeddings for retrieval and semantic search. Successful OpenRouter requests and embeddings may be retained and used to train...
Specifications
- Model ID
- liquid/lfm-2.5-embedding-350m:free
- Canonical ID
- liquid/lfm-2.5-embedding-350m-20260818
- Creator
- liquid
- Added to OpenRouter
- 18.08.2026
- Context length
- 512
- Modality
- text->embeddings
- Input modalities
- text
- Output modalities
- embeddings
- Token pricing
- USD per 1M tokens: in $0 / out $0
- Knowledge cutoff
- —
- Expiration date
- 13.11.2026
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | — |
| Coding | No measurement | — |
| Agents | No measurement | — |
| Writing | No measurement | — |
| Reasoning | No measurement | — |
| Document analysis | No measurement | — |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | output: embeddings |
Evaluations
No evaluations are supplied for this model.
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
Provider details are unavailable right now; the catalog data above is unaffected.
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.