Knowledge · Benchmark profile
A display-only multilingual QA retrieval benchmark reported by Liquid AI for LFM2.5 retriever models, using Recall@20 across 11 languages.
Data verified 21 Jul 2026 · Methodology 1.6.0
Benchmark score on MKQA-11 multilingual retrieval
LFM2.5-ColBERT-350M leads at 69.4%, followed by LFM2.5-Embedding-350M (69.1%).
Visual analysis
Switch between model placement, score distribution and descriptive provider averages. Every view uses the same sourced leaderboard.
2 ranked models · higher is better
| Model | Rank | Provider | Score |
|---|---|---|---|
| LFM2.5-ColBERT-350M | #1 | LiquidAI | 69.4% |
| LFM2.5-Embedding-350M | #2 | LiquidAI | 69.1% |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | LFM2.5-ColBERT-350M lfm2-5-colbert-350m | LiquidAI | open | Estimated reference | 69.4% |
| #2 | LFM2.5-Embedding-350M lfm2-5-embedding-350m | LiquidAI | open | Estimated reference | 69.1% |
About MKQA-11 multilingual retrieval
A display-only multilingual QA retrieval benchmark reported by Liquid AI for LFM2.5 retriever models, using Recall@20 across 11 languages. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
A display-only multilingual QA retrieval benchmark reported by Liquid AI for LFM2.5 retriever models, using Recall@20 across 11 languages.
LFM2.5-ColBERT-350M by LiquidAI currently leads with 69.4%.
2 models in the LuminaBench cohort have a qualifying score on this benchmark.
Related
2026 · 10 results · Reference
Artificial Analysis MMLU-Pro2026 · 4 results · Reference
Massive Multitask Language Understanding2020 · 8 results · Reference
Graduate-Level Google-Proof Q&A2023 · 70 results · Reference
SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines2025 · 19 results · Reference
Massive Multitask Language Understanding Professional2024 · 42 results · Reference
