# Text Embedding on MTEB (Quality and Efficiency)

Source: [Text Embedding on MTEB (Quality and Efficiency)](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-10T09:59:19.336Z

Evaluation of text embedding models using the Massive Text Embedding Benchmark (MTEB) suite, reporting both standard quality metrics and operational performance indicators such as latency and throughput.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Text Embedding |
| Dataset (including any stated split) | MTEB |
| Metric | MTEB Quality Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [MTEB Quality Score](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency.md?metric=MTEB%20Quality%20Score>): Higher is better; unit: SCALAR.

- [Model Size (MB)](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency.md?metric=Model%20Size%20(MB)>): Lower is better; unit: SIZE.

- [p50 Latency (ms)](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency.md?metric=p50%20Latency%20(ms)>): Lower is better; unit: TIME.

- [p99 Latency (ms)](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency.md?metric=p99%20Latency%20(ms)>): Lower is better; unit: TIME.

- [Throughput (RPS)](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency.md?metric=Throughput%20(RPS)>): Higher is better; unit: SCALAR.

- [Memory Usage (GB)](<https://www.sota2.com/research/sota/text-embedding-on-mteb-quality-and-efficiency.md?metric=Memory%20Usage%20(GB)>): Lower is better; unit: SIZE.

## Recorded results

| Method | MTEB Quality Score | Result date | Report paper |
| --- | --- | --- | --- |
| Sentence-BERT | 67.8 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| TensorRT-BERT | 67.8 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| Quantized BERT | 65.8 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| GTE-tiny | 65.2 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| BGE-micro | 64.3 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| SwiftEmbed | 60.6 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| FastText | 50.4 | Not recorded | 6902f0f13ef7ccd74e5e4258 |
| GloVe-840B | 48.5 | Not recorded | 6902f0f13ef7ccd74e5e4258 |

### Sentence-BERT

MTEB Quality Score: 67.8. Result ID: 69afeb6d6dbb7075b958cbf7.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ecf1e209131cea424f7.md>)

[Implementation](<https://github.com/embeddings-benchmark/mteb>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### TensorRT-BERT

MTEB Quality Score: 67.8. Result ID: 69afeb6d6dbb7075b958cbfa.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### Quantized BERT

MTEB Quality Score: 65.8. Result ID: 69afeb6d6dbb7075b958cbfb.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e9dc.md>)

[Implementation](<https://github.com/NervanaSystems/nlp-architect>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### GTE-tiny

MTEB Quality Score: 65.2. Result ID: 69afeb6d6dbb7075b958cbf8.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ecf1e209131cea424f7.md>)

[Implementation](<https://github.com/embeddings-benchmark/mteb>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### BGE-micro

MTEB Quality Score: 64.3. Result ID: 69afeb6d6dbb7075b958cbf9.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ecf1e209131cea424f7.md>)

[Implementation](<https://github.com/embeddings-benchmark/mteb>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### SwiftEmbed

MTEB Quality Score: 60.6. Result ID: 69afeb6d6dbb7075b958cbfe.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### FastText

MTEB Quality Score: 50.4. Result ID: 69afeb6d6dbb7075b958cbfc.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ecf1e209131cea424f7.md>)

[Implementation](<https://github.com/embeddings-benchmark/mteb>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

### GloVe-840B

MTEB Quality Score: 48.5. Result ID: 69afeb6d6dbb7075b958cbfd.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f13ef7ccd74e5e4258.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ecf1e209131cea424f7.md>)

[Implementation](<https://github.com/embeddings-benchmark/mteb>)

Recorded evaluation configuration: {"Hardware":"AMD EPYC 7742"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
