Model type
Embedding model
6 Downloads
Capabilities
Minimum system memory
Tags
Last updated
Updated 12 hours agobyModel type
Embedding model
6 Downloads
Capabilities
Minimum system memory
Tags
Last updated
Updated 12 hours agobyREADME
Sources
The underlying model files this model uses
README
Sources
The underlying model files this model uses
Google's EmbeddingGemma 2 for search and retrieval on a CPU: it turns text into 768-dimensional vectors.
embeddinggemma-2-Artemis-Q4_K_S.gguf.TurinTech models, built with Artemis. GGUF files of google/embeddinggemma-2 for text embeddings.
Start with embeddinggemma-2-Artemis-Q4_K_S.gguf. It returns 768-dimensional vectors through LM Studio's /v1/embeddings endpoint. It is not a chat model.
lms get turintech/embeddinggemma-2-artemis@q4_k_s --gguf
You can also paste turintech/embeddinggemma-2-Artemis-GGUF into LM Studio's Discover search and pick that exact file.
Checked on October 25, 2026 with LM Studio 0.4.26 (headless llmster 0.0.25-1) and its CPU llama.cpp engine 2.64.0: the definition loaded the file above and /v1/embeddings returned 768-dimensional vectors. The source card's retrieval numbers were measured with llama.cpp and Ollama; check retrieval quality in your own setup, since pooling and normalisation settings affect results.
| File | Download GB | SHA-256 |
|---|---|---|
| embeddinggemma-2-Artemis-Q4_K_S.gguf | 0.157 | f76351ba13988274750620577c88911b195a15321c37fc991bed485ed8f81f45 |
| embeddinggemma-2-Artemis-Q5_K_S.gguf | 0.182 | 476a432281b3c9c8f16b5286bbe1aa482357e4b15fe69c3f8008ade36fd8f381 |
See the source card for comparisons, limitations and file-specific provenance. Smaller files can give different answers. This is an independent TurinTech conversion, not affiliated with or endorsed by Google.
Our two files against Unsloth's Q4 file and the original, on 96 queries and 1,024 documents from SciFact and FiQA. This development panel was used to choose the files.
| File | Size | nDCG@10 | Recall@10 |
|---|---|---|---|
| Ours, Q4_K_S (default) | 156.935 MB | 0.859481 | 0.914856 |
| Ours, Q5_K_S | 182.036 MB | 0.872079 | 0.928224 |
| Unsloth Q4 | 175.674 MB | 0.857125 | 0.920933 |
| Original FP32 | n/a | 0.856404 | 0.928224 |
The default is the smallest file and gives up a little Recall@10 against Unsloth's Q4. Peak RAM and GPU compatibility are not established yet. How it was tested.
Every file we publish for this model. Sizes are download sizes, not memory.
| File | Size | Use it for |
|---|---|---|
embeddinggemma-2-Artemis-Q4_K_S.gguf | 157 MB | Default: smaller download; measured Recall below Unsloth Q4 |
embeddinggemma-2-Artemis-Q5_K_S.gguf | 182 MB | Larger file with higher observed nDCG and Recall on the exposed development panel |
-hf turintech/embeddinggemma-2-Artemis-GGUF:Q4_K_S.turintech/embeddinggemma-2-artemis offers every model file above.Smaller weights do not shrink the embedding index.
License: Apache-2.0. Keep the source repository's LICENSE and NOTICE with redistributed weights.
Based on
GGUF
Google's EmbeddingGemma 2 for search and retrieval on a CPU: it turns text into 768-dimensional vectors.
embeddinggemma-2-Artemis-Q4_K_S.gguf.TurinTech models, built with Artemis. GGUF files of google/embeddinggemma-2 for text embeddings.
Start with embeddinggemma-2-Artemis-Q4_K_S.gguf. It returns 768-dimensional vectors through LM Studio's /v1/embeddings endpoint. It is not a chat model.
lms get turintech/embeddinggemma-2-artemis@q4_k_s --gguf
You can also paste turintech/embeddinggemma-2-Artemis-GGUF into LM Studio's Discover search and pick that exact file.
Checked on October 25, 2026 with LM Studio 0.4.26 (headless llmster 0.0.25-1) and its CPU llama.cpp engine 2.64.0: the definition loaded the file above and /v1/embeddings returned 768-dimensional vectors. The source card's retrieval numbers were measured with llama.cpp and Ollama; check retrieval quality in your own setup, since pooling and normalisation settings affect results.
| File | Download GB | SHA-256 |
|---|---|---|
| embeddinggemma-2-Artemis-Q4_K_S.gguf | 0.157 | f76351ba13988274750620577c88911b195a15321c37fc991bed485ed8f81f45 |
| embeddinggemma-2-Artemis-Q5_K_S.gguf | 0.182 | 476a432281b3c9c8f16b5286bbe1aa482357e4b15fe69c3f8008ade36fd8f381 |
See the source card for comparisons, limitations and file-specific provenance. Smaller files can give different answers. This is an independent TurinTech conversion, not affiliated with or endorsed by Google.
Our two files against Unsloth's Q4 file and the original, on 96 queries and 1,024 documents from SciFact and FiQA. This development panel was used to choose the files.
| File | Size | nDCG@10 | Recall@10 |
|---|---|---|---|
| Ours, Q4_K_S (default) | 156.935 MB | 0.859481 | 0.914856 |
| Ours, Q5_K_S | 182.036 MB | 0.872079 | 0.928224 |
| Unsloth Q4 | 175.674 MB | 0.857125 | 0.920933 |
| Original FP32 | n/a | 0.856404 | 0.928224 |
The default is the smallest file and gives up a little Recall@10 against Unsloth's Q4. Peak RAM and GPU compatibility are not established yet. How it was tested.
Every file we publish for this model. Sizes are download sizes, not memory.
| File | Size | Use it for |
|---|---|---|
embeddinggemma-2-Artemis-Q4_K_S.gguf | 157 MB | Default: smaller download; measured Recall below Unsloth Q4 |
embeddinggemma-2-Artemis-Q5_K_S.gguf | 182 MB | Larger file with higher observed nDCG and Recall on the exposed development panel |
-hf turintech/embeddinggemma-2-Artemis-GGUF:Q4_K_S.turintech/embeddinggemma-2-artemis offers every model file above.Smaller weights do not shrink the embedding index.
License: Apache-2.0. Keep the source repository's LICENSE and NOTICE with redistributed weights.
Based on
GGUF
curl http://localhost:1234/v1/embeddings -H 'Content-Type: application/json' \
-d '{"model": "turintech/embeddinggemma-2-artemis", "input": "How do I reset my password?"}'
curl http://localhost:1234/v1/embeddings -H 'Content-Type: application/json' \
-d '{"model": "turintech/embeddinggemma-2-artemis", "input": "How do I reset my password?"}'