Model results

MTEB

Choose a shared evaluation protocol to compare model scores. Versions, sources and configurations stay separate.

Atlas snapshot captured 10/19/2026, 10:20:28 AM UTC. This is an import timestamp, not the evaluation date. Coverage may be incomplete.

Reviewed recommendation and limits

Start with retrieval tasks matching your language and domain. Inspect RTEB held-out subsets alongside public-task results when assessing transfer to unseen data.

Select and report the benchmark edition, retrieval task set, language, split and model settings. Preserve task metrics and aggregation. Keep MTEB English v2 and each RTEB open/private edition separate.. Imported groups below may use different protocols; this review does not certify each run. Public-task scores can reflect training exposure. Private RTEB data limits independent reproduction, and its launch report flags QA-derived keyword overlap. Imported English v2 results do not represent RTEB or complete RAG answer quality.

Publisher results MTEB task selection RTEB design and limitations MTEB research and attribution

← All evaluation groups

MTEB English retrieval

Source: MTEB
Version: v2
Measurement date: Not reported
Metric: % · Higher is better
Evaluation scope: MTEB English v2; mean nDCG@10 across 10/10 retrieval tasks: ArguAna, ClimateFEVERHardNegatives, CQADupstackGamingRetrieval, CQADupstackUnixRetrieval, FEVERHardNegatives, FiQA2018, HotpotQAHardNegatives, SCIDOCS, Touche2020Retrieval.v3, TRECCOVID; source-reported training overlap: FEVERHardNegatives, HotpotQAHardNegatives; experiment configuration: not supplied; result date not supplied

Positions are withheld because at least one result lacks explicit independent provenance or a documented evaluation scope. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

Qwen3 Embedding 8B
qwen/qwen3-embedding-8b
69.4%
View modelEvidence aggregated
Qwen3 Embedding 4B
qwen/qwen3-embedding-4b
68.5%
View modelEvidence aggregated
Gemini Embedding 001
google/gemini-embedding-001
64.4%
View modelEvidence aggregated
Multilingual-E5-Large
intfloat/multilingual-e5-large-20251117
51.5%
View modelEvidence aggregated