RTEB
Qwen/Qwen3-Embedding-8B 69.4% tops its leaderboard
Not in model cards Embedding models retrieve documents on open and private datasets across 20 languages and domains such as law, healthcare, code and finance. Scored by NDCG@10.
MTEB maintainers run the private sets and publish only statistics and samples, so a drop from open to private scores flags overfitting. Shown as the Retrieval section of the MTEB board.
- Made by
- MTEB community and Hugging Face (with MongoDB)
- Released
- October 2025
- Kind
- Other
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- Public, plus a private test set
11
Entries on its own leaderboard, 0 models we track
Qwen/Qwen3-Embedding-8B tops the leaderboard with 69.4%
Best score per entry from the benchmark's own leaderboard, read on November 12th, 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.
- 69.4 Qwen/Qwen3-Embedding-8B
- 67.1 Qwen/Qwen3-Embedding-4B
- 52.5 hotchpotch/bekko-embedding-v1-a25m
- 51.5 microsoft/harrier-oss-v1-270m
- 48.6 hotchpotch/bekko-embedding-v1-a8m
- 46.7 codefuse-ai/F2LLM-v2-80M
- 46.1 nomic-ai/nomic-embed-text-v2-moe
- 42.3 intfloat/multilingual-e5-small
- 34.9 mteb/baseline-bm25s
- 23.3 sentence-transformers/static-similarity-mrl-multilingual-v1
- 22.0 minishlab/potion-multilingual-128M
- 1 Qwen/Qwen3-Embedding-8B RTEB beta; all tasks 69.4%
- 2 Qwen/Qwen3-Embedding-4B RTEB beta; all tasks 67.1%
- 3 hotchpotch/bekko-embedding-v1-a25m RTEB beta; all tasks 52.5%
- 4 microsoft/harrier-oss-v1-270m RTEB beta; all tasks 51.5%
- 5 hotchpotch/bekko-embedding-v1-a8m RTEB beta; all tasks 48.6%
- 6 codefuse-ai/F2LLM-v2-80M RTEB beta; all tasks 46.7%
- 7 nomic-ai/nomic-embed-text-v2-moe RTEB beta; all tasks 46.1%
- 8 intfloat/multilingual-e5-small RTEB beta; all tasks 42.3%
- 9 mteb/baseline-bm25s RTEB beta; all tasks 34.9%
- 10 sentence-transformers/static-similarity-mrl-multilingual-v1 RTEB beta; all tasks 23.3%
- 11 minishlab/potion-multilingual-128M RTEB beta; all tasks 22%
Source: Official RTEB leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Qwen/Qwen3-Embedding-8B | 69.4% | RTEB beta; all tasks | Official leaderboard | |
| 2 | Qwen/Qwen3-Embedding-4B | 67.1% | RTEB beta; all tasks | Official leaderboard | |
| 3 | hotchpotch/bekko-embedding-v1-a25m | 52.5% | RTEB beta; all tasks | Official leaderboard | |
| 4 | microsoft/harrier-oss-v1-270m | 51.5% | RTEB beta; all tasks | Official leaderboard | |
| 5 | hotchpotch/bekko-embedding-v1-a8m | 48.6% | RTEB beta; all tasks | Official leaderboard | |
| 6 | codefuse-ai/F2LLM-v2-80M | 46.7% | RTEB beta; all tasks | Official leaderboard | |
| 7 | nomic-ai/nomic-embed-text-v2-moe | 46.1% | RTEB beta; all tasks | Official leaderboard | |
| 8 | intfloat/multilingual-e5-small | 42.3% | RTEB beta; all tasks | Official leaderboard | |
| 9 | mteb/baseline-bm25s | 34.9% | RTEB beta; all tasks | Official leaderboard | |
| 10 | sentence-transformers/static-similarity-mrl-multilingual-v1 | 23.3% | RTEB beta; all tasks | Official leaderboard | |
| 11 | minishlab/potion-multilingual-128M | 22% | RTEB beta; all tasks | Official leaderboard |
No lab model card reports it yet
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.
More in Other
- Personal Superintelligence Index (Meta) In use Meta's composite of seven personal and knowledge-work evals: Agents' Last Exam, AutomationBench-AA, FocusBench, JobBench, GDP.pdf, VBT and DRACO.
- Artificial Analysis Intelligence Index v5 In use Version 5 of Artificial Analysis's index: nine evaluations in four equal categories, most with private test sets, run three times each in its own agent harness.
Gemini 4 Carbon 59.0 - SAGE v1.1 (Vals) In use Models grade handwritten undergraduate math and computer science work against rubrics. Scored by class-balanced accuracy of each rubric check against expert grading. Claude Opus 5.5 91.3%
- MTEB Fading Massive Text Embedding Benchmark: retrieval, clustering, classification across languages and modalities.
sonar 42% - Anthropic ECI (AECI) Only in system cards Index that combines a basket of capability benchmarks into one capability score to track trends across Anthropic models. Claude Opus 5.5 169.4
- SPEED-Bench Only in system cards Measures speculative decoding across varied prompts and domains. Reports the average acceptance length of draft tokens, so higher means faster generation at the same output.
Nemotron 3 Ultra 5.15