Quick start
ollama run phonon:latestRunnable variants
Memory is a rough file-size estimate; context length, runtime and GPU offload change it.
12 of 14 matching variants shown.14 of 14 variants shown.
| Tag | Type | Inputs | Digest | Download | Parameters | Quantization | Context | Est. memory | Run |
|---|---|---|---|---|---|---|---|---|---|
| 31m-en | Typelocal | Inputstext | Digest6667dd188d2e | Download104 MB | Parameters31m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:31m-en |
| 40m-de | Typelocal | Inputstext | Digest6ef2680491eb | Download123 MB | Parameters40m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:40m-de |
| 40m-en | Typelocal | Inputstext | Digestea44df6fb977 | Download123 MB | Parameters40m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:40m-en |
| 40m-es | Typelocal | Inputstext | Digestd5df2ae8bc2d | Download123 MB | Parameters40m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:40m-es |
| 40m-fr | Typelocal | Inputstext | Digestda29147f2cba | Download123 MB | Parameters40m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:40m-fr |
| 40m-multilingual | Typelocal | Inputstext | Digesteda44c0f4703 | Download123 MB | Parameters40m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:40m-multilingual |
| 40m-pt | Typelocal | Inputstext | Digeste8fb87dea0cd | Download123 MB | Parameters40m | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:40m-pt |
| de | Typelocal | Inputstext | Digest6ef2680491eb | Download123 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:de |
| en | Typelocal | Inputstext | Digest6667dd188d2e | Download104 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:en |
| es | Typelocal | Inputstext | Digestd5df2ae8bc2d | Download123 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:es |
| fr | Typelocal | Inputstext | Digestda29147f2cba | Download123 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:fr |
| latestdefault | Typelocal | Inputstext | Digest6667dd188d2e | Download104 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:latest |
| multilingual | Typelocal | Inputstext | Digesteda44c0f4703 | Download123 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:multilingual |
| pt | Typelocal | Inputstext | Digeste8fb87dea0cd | Download123 MB | ParametersUnknown | QuantizationUnknown | Context8K | Est. memory≈ 1.5 GB | ollama run phonon:pt |
Related models
SmolLM2 is a family of compact language models available in three size: 135M, 360M, and 1.7B parameters.
🪐 A family of small models with 135M, 360M, and 1.7B parameters, trained on a new high-quality dataset.
The IBM Granite Embedding 30M and 278M models models are text-only dense biencoder embedding models, with 30M available in English only and 278M serving multilingual use cases.
Embedding models on very large sentence level datasets.