Mixedbread Quality Evals
How Mixedbread performs on public retrieval and agent benchmarks, measured against the strongest competing models.
Every result comes from a public benchmark on its standard split and evaluation script. Retrieval numbers are Wholembed V3 served through the Stores API; agent numbers come from Toast 1 or third-party agents running on Mixedbread Search, with cost and latency measured per query at list prices. Each benchmark lists its dataset, exact methodology and testing dates.
How we measure
Every result comes from a public benchmark on its standard split and evaluation script. Retrieval numbers are Wholembed V3 served through the Stores API; agent numbers come from Toast 1 or third-party agents running on Mixedbread Search, with cost and latency measured per query at list prices. Each benchmark lists its dataset, exact methodology and testing dates.
Wholembed V3, served through the Stores API, against other embedding models on public retrieval evals.
Video
Text-to-video retrieval (general clips, localized moments, instructional).
Per-dataset breakdown
| Model | MSRVTT Open MSRVTT dataset | DiDemo Open DiDemo dataset | YouCook2 Open YouCook2 dataset |
|---|---|---|---|
| Mixedbread | 64.4 | 72.3 | 64.2 |
| Gemini Embedding 2 | 67.6 | 68.1 | 52.5 |
| Voyage Multimodal 3.5 | 63.3 | 66.8 | 34.0 |
| Google Multimodal 001 | 58.3 | 51.0 | 36.1 |
NDCG@10 · higher is betterTested March 2026
Dataset and methodology
MSRVTT (general web video clips with text descriptions), DiDemo (localized text-to-moment retrieval), YouCook2 (instructional cooking videos with stepwise captions). Reported numbers are the across-dataset average and per-dataset NDCG@10.
We evaluate retrieval quality using publicly available eval datasets. Each model encodes the corpus and queries using its standard inference pipeline, then we compute the reported metric over the full test split. We report results as published by the eval or as measured on identical splits with default evaluation scripts.