LLMTR / llmtr/embeddinggemma-300m
EmbeddingGemma 300M - access through LLMTR
EmbeddingGemma 300M is a Turkey-hosted embedding (vector) model (Google EmbeddingGemma 300M). It does not generate text; it converts your input text into a 768-dimensional numeric vector for semantic search, RAG, similarity comparison, classification, and clustering. It is invoked through the OpenAI-compatible /v1/embeddings schema; batch input (an array in the `input` field) is supported and recommended. Vector dimensions can be MRL-truncated to 512/256/128. It accepts up to 2048 tokens per text and supports 100+ languages including Turkish. The same text always produces the same vector (deterministic). For best search/RAG quality, prefix queries with `task: search result | query: ...` and documents with `title: none | text: ...`.
Technical specifications
| Canonical ID | llmtr/embeddinggemma-300m |
|---|---|
| Provider | LLMTR |
| Context window | 2,048 tokens |
| Operations | EMBEDDINGS |
| Modalities | text, embedding |
Pricing
A 6% platform margin applies to credit top-ups; model usage prices are not separately marked up.
| Operation | Metric | Unit | Price |
|---|---|---|---|
| EMBEDDINGS | INPUT_TEXT | PER_1M_TOKENS | $0.100000 |
Example usage
With existing OpenAI SDK flows, change only the base URL and model identifier.
curl https://llmtr.com/v1/chat/completions \
-H "Authorization: Bearer llmtr-your_key" \
-H "Content-Type: application/json" \
-d '{"model":"llmtr/embeddinggemma-300m","messages":[{"role":"user","content":"Hello"}]}'Related models
- Gemma 4 llmtr/gemma-4
- Qwen 3.6 35B-A3B llmtr/qwen3-6-35b
- MedGemma 4B llmtr/medgemma-4b
- Trendyol 7B llmtr/trendyol-7b
- Magibu 11B v8 llmtr/magibu-11b-v8
- Sincap llmtr/sincap
- GPT-5.5 openai/gpt-5.5
- GPT-5.4 openai/gpt-5.4