TalentCLEF 2025 Multilingual Job Title Matching
Models rank job titles by how closely they match a query title within or across languages. Organizers score mean average precision.
The official results include multilingual E5 and Llama 3.3 runs. AlexU-NLP scored 0.534 mean average precision across English, Spanish, and German.
- Released
- September 2024
- Kind
- Other professions
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- Public, plus a private test set
- Measures
- Capability
- Subject
- Economics and business
- Interaction
- Single turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- A host
- In views
- Law and finance
multilingual-e5-large-instruct tops the leaderboard with 0.53
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 0.53 multilingual-e5-large-instruct
- 0.52 paraphrase-multilingual-mpnet-base-v2
- 0.52 Alibaba-NLP/gte-Qwen2-7B-instruct
- 0.48 ntfloat/multilingual-e5-large
- 0.48 BAAI/bge-m3
- 0.48 Alibaba-NLP/gte-multilingual-base
- 0.47 Linq-AI-Research/Linq-Embed-Mistral
- 0.46 multilingual-e5-large
- 0.44 intfloat/multilingual-e5-large-instruct
- 0.42 snowflake-arctic-embed-l-v2.0
- 0.41 MPNET base v2+DistilUSE multilingual cased v1
- 0.41 BAAI/bge-small-en-v1.5
- 0.40 MPNet
- 0.36 MiniLM
- 0.36 paraphrase-multilingual-MiniLM-L12-v2
- 0.18 Llama 3.3
- 1 multilingual-e5-large-instruct 283782, Retrieval, Contrastive Learning, InfoNCE 0.53
- 2 paraphrase-multilingual-mpnet-base-v2 284991, Retrieval, Contrastive Learning, InfoNCE 0.52
- 3 Alibaba-NLP/gte-Qwen2-7B-instruct 275330, Retrieval, Contrastive Learning, CachedGIST 0.52
- 4 ntfloat/multilingual-e5-large 276289, Retrieval, Contrastive Learning, InfoNCE 0.48
- 5 BAAI/bge-m3 275315, Retrieval, Contrastive Learning, GIST 0.48
- 6 Alibaba-NLP/gte-multilingual-base 275317, Retrieval, Contrastive Learning, GIST 0.48
- 7 Linq-AI-Research/Linq-Embed-Mistral 276155, Retrieval, Zero-shot 0.47
- 8 multilingual-e5-large + text-embedding-3-large + gemini-embedding-exp-03-07 284052, Retrieval, Zero-shot 0.46
- 9 intfloat/multilingual-e5-large-instruct 279236, Retrieval, Classification, CrossEntropy 0.44
- 10 snowflake-arctic-embed-l-v2.0 280184, Retrieval, Contrastive Learning, InfoNCE 0.42
- 11 MPNET base v2+DistilUSE multilingual cased v1 276052, Retrieval, Ensemble 0.41
- 12 BAAI/bge-small-en-v1.5 275311, Retrieval, Contrastive Learning, GIST 0.41
- 13 MPNet 285019, Retrieval, Zero-shot 0.40
- 14 MiniLM 284612, Retrieval, Contrastive Learning 0.36
- 15 paraphrase-multilingual-MiniLM-L12-v2 Retrieval, Zero-shot 0.36
- 16 Llama 3.3 285265, Retrieval, Zero-shot 0.18
Source: Official TalentCLEF 2025 Multilingual Job Title Matching leaderboard
| # | Entry | Score | Setting | Submitted by | Date | Source |
|---|---|---|---|---|---|---|
| 1 | multilingual-e5-large-instruct 0.6B | 0.53 | 283782, Retrieval, Contrastive Learning, InfoNCE, AlexU-NLP | AlexU-NLP | Official leaderboard | |
| 2 | paraphrase-multilingual-mpnet-base-v2 0.3B | 0.52 | 284991, Retrieval, Contrastive Learning, InfoNCE, TechWolf | TechWolf | Official leaderboard | |
| 3 | Alibaba-NLP/gte-Qwen2-7B-instruct 7B | 0.52 | 275330, Retrieval, Contrastive Learning, CachedGIST, pjmathematician | pjmathematician | Official leaderboard | |
| 4 | ntfloat/multilingual-e5-large 0.6B | 0.48 | 276289, Retrieval, Contrastive Learning, InfoNCE, NLPnorth | NLPnorth | Official leaderboard | |
| 5 | BAAI/bge-m3 0.6B | 0.48 | 275315, Retrieval, Contrastive Learning, GIST, pjmathematician | pjmathematician | Official leaderboard | |
| 6 | Alibaba-NLP/gte-multilingual-base 0.3B | 0.48 | 275317, Retrieval, Contrastive Learning, GIST, pjmathematician | pjmathematician | Official leaderboard | |
| 7 | Linq-AI-Research/Linq-Embed-Mistral 7.1B | 0.47 | 276155, Retrieval, Zero-shot, NLPnorth | NLPnorth | Official leaderboard | |
| 8 | multilingual-e5-large + text-embedding-3-large + gemini-embedding-exp-03-07 0.6B | 0.46 | 284052, Retrieval, Zero-shot, NT | NT | Official leaderboard | |
| 9 | intfloat/multilingual-e5-large-instruct 0.6B | 0.44 | 279236, Retrieval, Classification, CrossEntropy, NLPnorth | NLPnorth | Official leaderboard | |
| 10 | snowflake-arctic-embed-l-v2.0 0.6B | 0.42 | 280184, Retrieval, Contrastive Learning, InfoNCE, HULAT-UC3M | HULAT-UC3M | Official leaderboard | |
| 11 | MPNET base v2+DistilUSE multilingual cased v1 0.3B | 0.41 | 276052, Retrieval, Ensemble, UDII-UPM | UDII-UPM | Official leaderboard | |
| 12 | BAAI/bge-small-en-v1.5 0B | 0.41 | 275311, Retrieval, Contrastive Learning, GIST, pjmathematician | pjmathematician | Official leaderboard | |
| 13 | MPNet 0.3B | 0.40 | 285019, Retrieval, Zero-shot, DS@GT - TalentCLEF | DS@GT - TalentCLEF | Official leaderboard | |
| 14 | MiniLM 0.1B | 0.36 | 284612, Retrieval, Contrastive Learning, VerbaNexAI | VerbaNexAI | Official leaderboard | |
| 15 | paraphrase-multilingual-MiniLM-L12-v2 0.1B | 0.36 | Retrieval, Zero-shot, TalentCLEF Baseline | TalentCLEF Baseline | Official leaderboard | |
| 16 | Llama 3.3 70B | 0.18 | 285265, Retrieval, Zero-shot, Ixa | Ixa | Official leaderboard |
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Law, finance and medicine
- MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy.
Qwen 3.8 Max 80.4%
- HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
- HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
- FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy.
Ling-3.1-flash 81.7%
- KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
- Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access Public tasks, plus a private test set.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the official leaderboard (16 entries).
- Pass Models scored 16 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on talentclef.github.io: 16 entries at the last read, none a model we track.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.