ModernBERT base msmarco

ModernBERT

Open WeightModel weights are publicly available — can be downloaded and self-hosted

Released December 20248K context≈ 6,144 words

A focused retrieval specialist that takes the ModernBERT architecture and fine-tunes it specifically for passage and document ranking tasks using the MS MbenchmarksO dataset. It produces dense embeddings optimized for semantic search and information retrieval rather than general-purpose text understanding. Its 8192-token context window gives it more room than older BERT-style models to handle longer passages.

Capabilities

Capability scores are AI-generated based on model documentation, benchmarks, and technical specifications. Learn more

Instruction Following

Moderate

Reasoning & Logic

ModernBERT base msmarco

ModernBERT

Open WeightModel weights are publicly available — can be downloaded and self-hosted

Released December 20248K context≈ 6,144 words

A focused retrieval specialist that takes the ModernBERT architecture and fine-tunes it specifically for passage and document ranking tasks using the MS MbenchmarksO dataset. It produces dense embeddings optimized for semantic search and information retrieval rather than general-purpose text understanding. Its 8192-token context window gives it more room than older BERT-style models to handle longer passages.

Capabilities

Capability scores are AI-generated based on model documentation, benchmarks, and technical specifications. Learn more

Instruction Following

Moderate

Reasoning & Logic

Glossary

ArchitectureThe underlying structural design of a neural network that defines how data flows through layers and components.BERTA foundational neural network architecture designed to understand the meaning of words in context by learning from large amounts of text.Context WindowThe maximum number of tokens a model can process in a single conversation or prompt.Dense EmbeddingsVector representations where most or all of the numbers contain meaningful information, as opposed to sparse embeddings where most numbers are zero.EmbeddingsNumerical representations of text that capture semantic meaning, allowing the model to measure similarity between different words or phrases.General-PurposeDesigned to handle a wide variety of different tasks rather than being specialized for one specific domain.Information RetrievalThe task of finding relevant documents or passages from a large collection in response to a user query.RankingThe process of ordering search results by relevance, determining which documents best match a user's query.RetrievalThe process of finding and returning relevant documents or information from a database based on a query.Semantic SearchA search method that finds results based on the meaning of text rather than just matching keywords, using embeddings to understand intent.TokenA small unit of text (a word, subword, or punctuation mark) that a language model breaks input into for processing.

Capabilities

Capabilities

Use Case Fit

Glossary