From 2691928d90f338f01aee6bf9c796e2425e37d5a1 Mon Sep 17 00:00:00 2001
From: Platform Bot <platform@dakera.ai>
Date: Wed, 24 Jun 2026 10:53:21 +0000
Subject: [PATCH] feat(sdk): add MODERNBERT_EMBED_BASE + GTE_MODERNBERT_BASE to
 EmbeddingModel (DAK-7098/7102)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 CHANGELOG.md           | 9 +++++++++
 pyproject.toml         | 2 +-
 src/dakera/__init__.py | 2 +-
 src/dakera/models.py   | 4 ++++
 4 files changed, 15 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 198a608..8b6a494 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,6 +7,15 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.12.7] - 2026-06-24
+
+### Added
+
+- **`EmbeddingModel.MODERNBERT_EMBED_BASE` / `EmbeddingModel.GTE_MODERNBERT_BASE`** — add two
+  new enum members matching server DAK-7098/DAK-7102: `"modernbert-embed-base"` (768d, MRL,
+  8192 tokens) and `"gte-modernbert-base"` (768d, MTEB retrieval 64.38). Pass via the `model`
+  parameter of `upsert_text()` and `query_text()`.
+
 ## [0.12.6] - 2026-06-23
 
 ### Fixed
diff --git a/pyproject.toml b/pyproject.toml
index da0f131..4f30f37 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "dakera"
-version = "0.12.6"
+version = "0.12.7"
 description = "Python SDK for Dakera - AI memory platform"
 readme = "README.md"
 license = {text = "MIT"}
diff --git a/src/dakera/__init__.py b/src/dakera/__init__.py
index 1204d45..c7bca80 100644
--- a/src/dakera/__init__.py
+++ b/src/dakera/__init__.py
@@ -158,7 +158,7 @@
         DakeraDelegationHelper,
     )
 
-__version__ = "0.12.6"
+__version__ = "0.12.7"
 __all__ = [
     # Clients
     "DakeraClient",
diff --git a/src/dakera/models.py b/src/dakera/models.py
index bdc0836..fc699ea 100644
--- a/src/dakera/models.py
+++ b/src/dakera/models.py
@@ -475,12 +475,16 @@ class EmbeddingModel(str, Enum):
     - MINILM: MiniLM-L6 - Fast, good quality (384 dimensions)
     - BGE_SMALL: BGE-small - Balanced performance (384 dimensions)
     - E5_SMALL: E5-small - High quality (384 dimensions)
+    - MODERNBERT_EMBED_BASE: ModernBERT-embed-base - 768 dimensions, MRL, 8192 tokens
+    - GTE_MODERNBERT_BASE: GTE-ModernBERT-base - 768 dimensions, MTEB retrieval 64.38
     """
 
     BGE_LARGE = "bge-large"
     MINILM = "minilm"
     BGE_SMALL = "bge-small"
     E5_SMALL = "e5-small"
+    MODERNBERT_EMBED_BASE = "modernbert-embed-base"
+    GTE_MODERNBERT_BASE = "gte-modernbert-base"
 
 
 @dataclass