kreuzberg 4.7.4 → 4.8.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3fffd5a1e2d066b0997155be101dade834002fd2805fd556b2e5b6b8a8d29be9
-  data.tar.gz: 0f1ce8406a8880327191fa4e49bdadd71f938ec47ecb3b7e38f4121008d16600
+  metadata.gz: c768f2adf392c1598da39e79b20dcc7f0c55774f2d3063c74c6ec72888742bac
+  data.tar.gz: 8a350998762668be79e7f4a3843812782f4b0812fdb41327d8d38f1f57c39073
 SHA512:
-  metadata.gz: f878aeea0ccb330d30f863707fc31ba0dce4a8b2eb5f6fca705e5d21337a68ea209328a8dc84ae565c20a996d3ae6f5e53149d570a5439f6746f27df3d1c5671
-  data.tar.gz: 1ca62afb51ffb9d85f3a820c7d44473d393bfdefab21f78723c52d3304082382e30d646c2751203d2f0a0d2ab346ec664856662ba7c212c565b35777fa8f167f
+  metadata.gz: 18b2dc0c39199bead5f1f310a38b5f288cba0d8ebb62d31505f13933a0af3bf9179f8377e7a8bd605c39d46e900869a4492f1bd829f853d6f1765263f4affd4d
+  data.tar.gz: 769b9c36d5f6722f97f3c3aebb97f1045382a6dbca96405e298c44bc268795d97781edb89d140c1bc2f0e1fb33c5efecb79c2e500469dd7a25fe2105c49081ba

data/README.md

@@ -22,7 +22,7 @@
     <img src="https://img.shields.io/maven-central/v/dev.kreuzberg/kreuzberg?label=Java&color=007ec6" alt="Java">
   </a>
   <a href="https://github.com/kreuzberg-dev/kreuzberg/releases">
-    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.7.4" alt="Go">
+    <img src="https://img.shields.io/github/v/tag/kreuzberg-dev/kreuzberg?label=Go&color=007ec6&filter=v4.8.0" alt="Go">
   </a>
   <a href="https://www.nuget.org/packages/Kreuzberg/">
     <img src="https://img.shields.io/nuget/v/Kreuzberg?label=C%23&color=007ec6" alt="C#">
@@ -42,7 +42,7 @@
 
   <!-- Project Info -->
   <a href="https://github.com/kreuzberg-dev/kreuzberg/blob/main/LICENSE">
-    <img src="https://img.shields.io/badge/License-MIT-007ec6" alt="License">
+    <img src="https://img.shields.io/badge/License-Elastic--2.0-blue.svg" alt="License">
   </a>
   <a href="https://docs.kreuzberg.dev">
     <img src="https://img.shields.io/badge/docs-kreuzberg.dev-007ec6" alt="Documentation">
@@ -419,7 +419,7 @@ Contributions are welcome! See [Contributing Guide](https://github.com/kreuzberg
 
 ## License
 
-MIT License - see LICENSE file for details.
+Elastic License 2.0 (ELv2) - see [LICENSE](../../LICENSE) for details.
 
 ## Support
 

data/ext/kreuzberg_rb/native/Cargo.lock

@@ -1565,9 +1565,9 @@ dependencies = [
 
 [[package]]
 name = "fastrand"
-version = "2.4.0"
+version = "2.4.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a043dc74da1e37d6afe657061213aa6f425f855399a11d3463c6ecccc4dfda1f"
+checksum = "9f1f227452a390804cdb637b74a86990f2a7d7ba4b7d5693aac9b4dd6defd8d6"
 
 [[package]]
 name = "fax"
@@ -2792,7 +2792,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg"
-version = "4.7.3"
+version = "4.7.4"
 dependencies = [
  "ahash",
  "async-trait",
@@ -2829,12 +2829,14 @@ dependencies = [
  "kreuzberg-pdfium-render",
  "kreuzberg-tesseract",
  "libc",
+ "liter-llm",
  "log",
  "lopdf",
  "mail-parser",
  "memchr",
  "memmap2",
  "mime_guess",
+ "minijinja",
  "ndarray",
  "num_cpus",
  "once_cell",
@@ -2880,12 +2882,12 @@ dependencies = [
  "utoipa",
  "whatlang",
  "yake-rust",
- "zip 8.5.0",
+ "zip 8.5.1",
 ]
 
 [[package]]
 name = "kreuzberg-ffi"
-version = "4.7.3"
+version = "4.7.4"
 dependencies = [
  "ahash",
  "async-trait",
@@ -2901,7 +2903,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-paddle-ocr"
-version = "4.7.3"
+version = "4.7.4"
 dependencies = [
  "geo-clipper",
  "geo-types",
@@ -2915,7 +2917,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-pdfium-render"
-version = "4.7.3"
+version = "4.7.4"
 dependencies = [
  "bitflags",
  "bytemuck",
@@ -2938,7 +2940,7 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-rb"
-version = "4.7.3"
+version = "4.7.4"
 dependencies = [
  "async-trait",
  "html-to-markdown-rs",
@@ -2955,13 +2957,13 @@ dependencies = [
 
 [[package]]
 name = "kreuzberg-tesseract"
-version = "4.7.3"
+version = "4.7.4"
 dependencies = [
  "cc",
  "cmake",
  "reqwest",
  "thiserror 2.0.18",
- "zip 8.5.0",
+ "zip 8.5.1",
 ]
 
 [[package]]
@@ -3060,6 +3062,27 @@ version = "0.8.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "92daf443525c4cce67b150400bc2316076100ce0b3686209eb8cf3c31612e6f0"
 
+[[package]]
+name = "liter-llm"
+version = "1.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0405bbc5926db49a5f73a4f503d9cac19413416c95e2fd736b1bfa8ce9491660"
+dependencies = [
+ "base64 0.22.1",
+ "bytes",
+ "futures-core",
+ "memchr",
+ "pin-project-lite",
+ "reqwest",
+ "secrecy",
+ "serde",
+ "serde_json",
+ "thiserror 2.0.18",
+ "tokio",
+ "toml 1.1.2+spec-1.1.0",
+ "tracing",
+]
+
 [[package]]
 name = "litrs"
 version = "1.0.0"
@@ -3269,6 +3292,12 @@ dependencies = [
  "libc",
 ]
 
+[[package]]
+name = "memo-map"
+version = "0.3.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "38d1115007560874e373613744c6fba374c17688327a71c1476d1a5954cc857b"
+
 [[package]]
 name = "mime"
 version = "0.3.17"
@@ -3285,6 +3314,16 @@ dependencies = [
  "unicase",
 ]
 
+[[package]]
+name = "minijinja"
+version = "2.19.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "805bfd7352166bae857ee569628b52bcd85a1cecf7810861ebceb1686b72b75d"
+dependencies = [
+ "memo-map",
+ "serde",
+]
+
 [[package]]
 name = "minimal-lexical"
 version = "0.2.1"
@@ -4391,6 +4430,7 @@ dependencies = [
  "hyper-util",
  "js-sys",
  "log",
+ "mime_guess",
  "native-tls",
  "percent-encoding",
  "pin-project-lite",
@@ -4398,16 +4438,21 @@ dependencies = [
  "rustls",
  "rustls-pki-types",
  "rustls-platform-verifier",
+ "serde",
+ "serde_json",
+ "serde_urlencoded",
  "sync_wrapper",
  "tokio",
  "tokio-native-tls",
  "tokio-rustls",
+ "tokio-util",
  "tower",
  "tower-http",
  "tower-service",
  "url",
  "wasm-bindgen",
  "wasm-bindgen-futures",
+ "wasm-streams",
  "web-sys",
 ]
 
@@ -4676,6 +4721,16 @@ version = "1.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "94143f37725109f92c262ed2cf5e59bce7498c01bcc1502d7b9afe439a4e9f49"
 
+[[package]]
+name = "secrecy"
+version = "0.10.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e891af845473308773346dc847b2c23ee78fe442e0472ac50e22a18a93d3ae5a"
+dependencies = [
+ "serde",
+ "zeroize",
+]
+
 [[package]]
 name = "security-framework"
 version = "3.7.0"
@@ -5382,6 +5437,7 @@ dependencies = [
  "bytes",
  "libc",
  "mio",
+ "parking_lot",
  "pin-project-lite",
  "signal-hook-registry",
  "socket2",
@@ -6098,6 +6154,19 @@ dependencies = [
  "wasmparser",
 ]
 
+[[package]]
+name = "wasm-streams"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9d1ec4f6517c9e11ae630e200b2b65d193279042e28edd4a2cda233e46670bbb"
+dependencies = [
+ "futures-util",
+ "js-sys",
+ "wasm-bindgen",
+ "wasm-bindgen-futures",
+ "web-sys",
+]
+
 [[package]]
 name = "wasmparser"
 version = "0.244.0"
@@ -6788,9 +6857,9 @@ dependencies = [
 
 [[package]]
 name = "zip"
-version = "8.5.0"
+version = "8.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2726508a48f38dceb22b35ecbbd2430efe34ff05c62bd3285f965d7911b33464"
+checksum = "dcab981e19633ebcf0b001ddd37dd802996098bc1864f90b7c5d970ce76c1d59"
 dependencies = [
  "crc32fast",
  "flate2",

data/ext/kreuzberg_rb/native/Cargo.toml

@@ -1,10 +1,10 @@
 [package]
 name = "kreuzberg-rb"
-version = "4.7.4"
+version = "4.8.0"
 edition = "2024"
 rust-version = "1.91"
 authors = ["Na'aman Hirschfeld <naaman@kreuzberg.dev>"]
-license = "MIT"
+license = "Elastic-2.0"
 repository = "https://github.com/kreuzberg-dev/kreuzberg"
 homepage = "https://kreuzberg.dev"
 documentation = "https://docs.rs/kreuzberg"

data/ext/kreuzberg_rb/native/src/config/types.rs

@@ -22,6 +22,7 @@ use kreuzberg::{
     ImageExtractionConfig, LanguageDetectionConfig, LayoutDetectionConfig, OcrConfig, OutputFormat, PdfConfig,
     PostProcessorConfig, TokenReductionConfig,
 };
+use kreuzberg::core::config::ContentFilterConfig;
 use magnus::value::ReprValue;
 use magnus::{Error, RArray, RHash, Ruby, TryConvert, Value};
 use std::fs;
@@ -50,6 +51,8 @@ pub fn parse_ocr_config(ruby: &Ruby, hash: RHash) -> Result<OcrConfig, Error> {
         auto_rotate: false,
         pipeline: None,
         quality_thresholds: None,
+        vlm_config: None,
+        vlm_prompt: None,
     };
 
     if let Some(val) = get_kw(ruby, hash, "tesseract_config")
@@ -860,6 +863,42 @@ pub fn parse_email_config(ruby: &Ruby, hash: RHash) -> Result<EmailConfig, Error
     Ok(config)
 }
 
+/// Parse ContentFilterConfig from Ruby Hash
+pub fn parse_content_filter_config(ruby: &Ruby, hash: RHash) -> Result<ContentFilterConfig, Error> {
+    let include_headers = if let Some(val) = get_kw(ruby, hash, "include_headers") {
+        bool::try_convert(val)?
+    } else {
+        false
+    };
+
+    let include_footers = if let Some(val) = get_kw(ruby, hash, "include_footers") {
+        bool::try_convert(val)?
+    } else {
+        false
+    };
+
+    let strip_repeating_text = if let Some(val) = get_kw(ruby, hash, "strip_repeating_text") {
+        bool::try_convert(val)?
+    } else {
+        true
+    };
+
+    let include_watermarks = if let Some(val) = get_kw(ruby, hash, "include_watermarks") {
+        bool::try_convert(val)?
+    } else {
+        false
+    };
+
+    let config = ContentFilterConfig {
+        include_headers,
+        include_footers,
+        strip_repeating_text,
+        include_watermarks,
+    };
+
+    Ok(config)
+}
+
 /// Parse ExtractionConfig from Ruby Hash
 pub fn parse_extraction_config(ruby: &Ruby, opts: Option<RHash>) -> Result<ExtractionConfig, Error> {
     let mut config = ExtractionConfig::default();
@@ -996,6 +1035,13 @@ pub fn parse_extraction_config(ruby: &Ruby, opts: Option<RHash>) -> Result<Extra
             config.email = Some(parse_email_config(ruby, email_hash)?);
         }
 
+        if let Some(val) = get_kw(ruby, hash, "content_filter")
+            && val.equal(ruby.qnil()).ok() != Some(true)
+        {
+            let content_filter_hash = RHash::try_convert(val)?;
+            config.content_filter = Some(parse_content_filter_config(ruby, content_filter_hash)?);
+        }
+
         if let Some(val) = get_kw(ruby, hash, "max_concurrent_extractions") {
             let value = usize::try_convert(val)?;
             config.max_concurrent_extractions = Some(value);

data/ext/kreuzberg_rb/native/src/embedding.rs

@@ -0,0 +1,87 @@
+//! Standalone embedding functions for Ruby.
+//!
+//! Exposes `embed_sync` and `embed` module functions that generate vector embeddings
+//! from a list of text strings using the configured ONNX model.
+
+use crate::error_handling::{kreuzberg_error, runtime_error};
+use crate::helpers::ruby_value_to_json;
+use magnus::{Error, RArray, RHash, Ruby, TryConvert, Value, scan_args::get_kwargs, scan_args::scan_args};
+use magnus::value::ReprValue;
+
+/// Parse an optional Ruby value (Hash or nil) into a `kreuzberg::EmbeddingConfig`.
+fn parse_embedding_config(ruby: &Ruby, config_val: Option<Value>) -> Result<kreuzberg::EmbeddingConfig, Error> {
+    match config_val {
+        None => Ok(Default::default()),
+        Some(val) => {
+            if val.equal(ruby.qnil())? {
+                return Ok(Default::default());
+            }
+            let json = ruby_value_to_json(val)?;
+            serde_json::from_value(json)
+                .map_err(|e| runtime_error(format!("Invalid embedding config: {}", e)))
+        }
+    }
+}
+
+/// Convert `Vec<Vec<f32>>` to a Ruby Array of Arrays of Floats.
+fn embeddings_to_ruby(ruby: &Ruby, embeddings: Vec<Vec<f32>>) -> Result<RArray, Error> {
+    let outer = ruby.ary_new_capa(embeddings.len());
+    for inner_vec in embeddings {
+        let inner = ruby.ary_new_capa(inner_vec.len());
+        for v in inner_vec {
+            inner.push(v as f64)?;
+        }
+        outer.push(inner)?;
+    }
+    Ok(outer)
+}
+
+/// Parse keyword args common to `embed_sync` and `embed`.
+/// Returns `(texts, config)`.
+fn parse_embed_args(
+    ruby: &Ruby,
+    args: &[Value],
+) -> Result<(Vec<String>, kreuzberg::EmbeddingConfig), Error> {
+    let parsed = scan_args::<(), (), (), (), RHash, ()>(args)?;
+    let kw = parsed.keywords;
+
+    let kw_args = get_kwargs::<_, (Value,), (Option<Value>,), ()>(kw, &["texts"], &["config"])?;
+    let (texts_val,) = kw_args.required;
+    let (config_opt,) = kw_args.optional;
+
+    let texts_arr = RArray::try_convert(texts_val)
+        .map_err(|_| runtime_error("texts must be an Array".to_string()))?;
+    let texts: Vec<String> = texts_arr
+        .into_iter()
+        .enumerate()
+        .map(|(i, v)| {
+            String::try_convert(v)
+                .map_err(|_| runtime_error(format!("texts[{}] must be a String", i)))
+        })
+        .collect::<Result<_, _>>()?;
+
+    let config = parse_embedding_config(ruby, config_opt)?;
+    Ok((texts, config))
+}
+
+/// Generate embeddings synchronously.
+///
+/// Keyword args: `texts:` (Array of String), `config:` (Hash, optional)
+/// Returns: Array of Arrays of Float (one per input text).
+pub fn embed_sync(args: &[Value]) -> Result<RArray, Error> {
+    let ruby = Ruby::get().expect("Ruby not initialized");
+    let (texts, config) = parse_embed_args(&ruby, args)?;
+    let embeddings = kreuzberg::embed_texts(&texts, &config).map_err(kreuzberg_error)?;
+    embeddings_to_ruby(&ruby, embeddings)
+}
+
+/// Generate embeddings (delegates to `embed_sync`).
+///
+/// Ruby's GVL prevents true async execution, so this simply delegates to
+/// the synchronous implementation to avoid creating a throwaway Tokio runtime.
+///
+/// Keyword args: `texts:` (Array of String), `config:` (Hash, optional)
+/// Returns: Array of Arrays of Float (one per input text).
+pub fn embed(args: &[Value]) -> Result<RArray, Error> {
+    embed_sync(args)
+}

data/ext/kreuzberg_rb/native/src/error_handling.rs

@@ -101,6 +101,13 @@ pub fn kreuzberg_error(err: KreuzbergError) -> Error {
                 )
             }
         }
+        KreuzbergError::Embedding { message, .. } => {
+            if let Some(class) = fetch_error_class("EmbeddingError") {
+                Error::new(class, message)
+            } else {
+                Error::new(ruby.exception_runtime_error(), format!("EmbeddingError: {}", message))
+            }
+        }
         other => Error::new(ruby.exception_runtime_error(), other.to_string()),
     }
 }

data/ext/kreuzberg_rb/native/src/lib.rs

@@ -6,6 +6,7 @@
 //! Provides extraction, OCR, chunking, and language detection for 30+ file formats.
 
 // Module declarations
+mod embedding;
 mod error_handling;
 mod gc_guarded_value;
 mod helpers;
@@ -457,6 +458,10 @@ fn init(ruby: &Ruby) -> Result<(), Error> {
     module.define_module_function("batch_extract_files", function!(batch_extract_files, -1))?;
     module.define_module_function("batch_extract_bytes", function!(batch_extract_bytes, -1))?;
 
+    // Embedding functions
+    module.define_module_function("embed_sync", function!(embedding::embed_sync, -1))?;
+    module.define_module_function("embed", function!(embedding::embed, -1))?;
+
     // PDF page iterator
     module.define_module_function("native_render_pdf_pages_iter", function!(render_pdf_pages_iter, 2))?;
     module.define_module_function("native_render_pdf_page", function!(native_render_pdf_page, 3))?;

data/ext/kreuzberg_rb/native/src/plugins/ocr_backend.rs

@@ -110,6 +110,7 @@ impl OcrBackend for RubyOcrBackend {
                 formatted_content: None,
                 uris: None,
                 ocr_internal_document: None,
+                structured_output: None,
             })
         })
     }

data/lib/kreuzberg/config.rb

@@ -856,6 +856,40 @@ module Kreuzberg
       end
     end
 
+    # Content filter configuration for controlling extraction of headers, footers,
+    # watermarks, and repeating text across document formats.
+    #
+    # @example Include headers and footers
+    #   filter = ContentFilter.new(include_headers: true, include_footers: true)
+    #
+    # @example Disable repeating text removal
+    #   filter = ContentFilter.new(strip_repeating_text: false)
+    #
+    class ContentFilter
+      attr_reader :include_headers, :include_footers, :strip_repeating_text, :include_watermarks
+
+      def initialize(
+        include_headers: false,
+        include_footers: false,
+        strip_repeating_text: true,
+        include_watermarks: false
+      )
+        @include_headers = include_headers ? true : false
+        @include_footers = include_footers ? true : false
+        @strip_repeating_text = strip_repeating_text ? true : false
+        @include_watermarks = include_watermarks ? true : false
+      end
+
+      def to_h
+        {
+          include_headers: @include_headers,
+          include_footers: @include_footers,
+          strip_repeating_text: @strip_repeating_text,
+          include_watermarks: @include_watermarks
+        }
+      end
+    end
+
     # Layout detection configuration
     #
     # @example Basic usage
@@ -951,7 +985,7 @@ module Kreuzberg
                   :max_concurrent_extractions, :output_format, :result_format,
                   :security_limits, :layout, :concurrency,
                   :cache_namespace, :cache_ttl_secs, :extraction_timeout_secs,
-                  :max_archive_depth, :acceleration, :email
+                  :max_archive_depth, :acceleration, :email, :content_filter
 
       # Alias for backward compatibility - image_extraction is the canonical name
       alias image_extraction images
@@ -977,7 +1011,7 @@ module Kreuzberg
         postprocessor token_reduction keywords html_options pages
         max_concurrent_extractions output_format result_format
         security_limits layout concurrency cache_namespace cache_ttl_secs extraction_timeout_secs
-        max_archive_depth acceleration email
+        max_archive_depth acceleration email content_filter
       ].freeze
 
       # Aliases for backward compatibility
@@ -1062,7 +1096,8 @@ module Kreuzberg
                      extraction_timeout_secs: nil,
                      max_archive_depth: 3,
                      acceleration: nil,
-                     email: nil)
+                     email: nil,
+                     content_filter: nil)
         kwargs = {
           use_cache: use_cache, enable_quality_processing: enable_quality_processing,
           force_ocr: force_ocr, disable_ocr: disable_ocr, force_ocr_pages: force_ocr_pages,
@@ -1080,7 +1115,8 @@ module Kreuzberg
           extraction_timeout_secs: extraction_timeout_secs,
           max_archive_depth: max_archive_depth,
           acceleration: acceleration,
-          email: email
+          email: email,
+          content_filter: content_filter
         }
         extracted = extract_from_hash(hash, kwargs)
 
@@ -1115,6 +1151,7 @@ module Kreuzberg
         @concurrency = normalize_config(params[:concurrency], Concurrency)
         @acceleration = normalize_config(params[:acceleration], Acceleration)
         @email = normalize_config(params[:email], Email)
+        @content_filter = normalize_config(params[:content_filter], ContentFilter)
         @max_concurrent_extractions = params[:max_concurrent_extractions]&.to_i
         @max_archive_depth = params[:max_archive_depth]&.to_i || 3
         @output_format = validate_output_format(params[:output_format])
@@ -1175,7 +1212,8 @@ module Kreuzberg
           token_reduction: @token_reduction&.to_h, keywords: @keywords&.to_h,
           html_options: @html_options&.to_h, pages: @pages&.to_h,
           layout: @layout&.to_h, concurrency: @concurrency&.to_h,
-          acceleration: @acceleration&.to_h, email: @email&.to_h
+          acceleration: @acceleration&.to_h, email: @email&.to_h,
+          content_filter: @content_filter&.to_h
         }
       end
 

data/lib/kreuzberg/errors.rb

@@ -11,6 +11,7 @@ module Kreuzberg
   ERROR_CODE_PARSING = 5
   ERROR_CODE_OCR = 6
   ERROR_CODE_MISSING_DEPENDENCY = 7
+  ERROR_CODE_EMBEDDING = 8
 
   module Errors
     class PanicContext
@@ -112,5 +113,8 @@ module Kreuzberg
 
     # Raised when an unsupported file format or MIME type is encountered
     class UnsupportedFormatError < Error; end
+
+    # Raised when embedding fails
+    class EmbeddingError < Error; end
   end
 end

data/lib/kreuzberg/extraction_api.rb

@@ -236,6 +236,41 @@ module Kreuzberg
       results
     end
 
+    # Asynchronously generate embeddings for multiple texts.
+    #
+    # Non-blocking embedding generation from a list of strings.
+    #
+    # @param texts [Array<String>] List of strings to embed.
+    # @param config [Config::Embedding, Hash, nil] Embedding configuration.
+    #
+    # @return [Array<Array<Float>>] Array of embedding vectors.
+    #
+    # @raise [Errors::EmbeddingError] If embedding generation fails.
+    #
+    # @example Generate embeddings asynchronously
+    #   texts = ["Hello, world!", "Kreuzberg is awesome."]
+    #   embeddings = Kreuzberg.embed(texts: texts)
+    #   puts embeddings.first.length # 384
+    def embed(texts:, config: nil)
+      opts = normalize_config(config)
+      native_embed(texts: texts.map(&:to_s), config: opts)
+    end
+
+    # Synchronously generate embeddings for multiple texts.
+    #
+    # Blocking embedding generation from a list of strings.
+    #
+    # @param texts [Array<String>] List of strings to embed.
+    # @param config [Config::Embedding, Hash, nil] Embedding configuration.
+    #
+    # @return [Array<Array<Float>>] Array of embedding vectors.
+    #
+    # @raise [Errors::EmbeddingError] If embedding generation fails.
+    def embed_sync(texts:, config: nil)
+      opts = normalize_config(config)
+      native_embed_sync(texts: texts.map(&:to_s), config: opts)
+    end
+
     # Synchronously extract content from multiple byte data sources.
     #
     # Processes multiple in-memory binary documents in a single batch operation. Results

data/lib/kreuzberg/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kreuzberg
-  VERSION = '4.7.4'
+  VERSION = '4.8.0'
 end

data/lib/kreuzberg.rb

@@ -59,10 +59,13 @@ module Kreuzberg
     alias native_batch_extract_bytes batch_extract_bytes
     alias native_clear_cache clear_cache
     alias native_cache_stats cache_stats
+    alias native_embed_sync embed_sync
+    alias native_embed embed
 
     private :native_extract_file_sync, :native_extract_bytes_sync, :native_batch_extract_files_sync
     private :native_extract_file, :native_extract_bytes, :native_batch_extract_files
     private :native_batch_extract_bytes_sync, :native_batch_extract_bytes
+    private :native_embed_sync, :native_embed
   end
 
   module_function :register_post_processor
@@ -94,6 +97,10 @@ module Kreuzberg
   module_function :validate_mime_type
 
   module_function :get_extensions_for_mime
+
+  module_function :embed_sync
+
+  module_function :embed
 end
 
 require_relative 'kreuzberg/cache_api'