Models
Which model should you pick?
A plain guide to the models you can run in Llamatik: how big they are, how much memory they need, and what each one is good for.
Start here
Gemma 3 270M (292 MB) answers quickly even on older phones and is multilingual.
Gemma 3 1B (806 MB) is a good balance. For more depth in many languages, try Qwen 3 1.7B (1.8 GB).
SmolVLM 500M (437 MB) describes images and reads the text in them. SmolVLM 256M (175 MB) is the lightest option.
Whisper Base q8_0 (82 MB) is a fast, accurate default. Whisper Small q8_0 (264 MB) copes better with accents and noise.
Chat
| Model | Download | Memory guide | Good for |
|---|---|---|---|
| Gemma 3 270M Instruct (Q8_0) | 292 MB | ~750 MB | FastestMultilingual |
| Gemma 3 1B Instruct (Q4_K_M) | 806 MB | ~1.6 GB | Balanced |
| Qwen 3 1.7B (Q8_0) | 1.8 GB | ~3.0 GB | Best hereMultilingual |
| Qwen 2.5 1.5B Instruct (Q2_K) | 753 MB | ~1.4 GB | MultilingualHeavily compressed |
| Llama 3.2 1B Instruct (Q2_K) | 581 MB | ~1.2 GB | MultilingualHeavily compressed |
| Phi-1.5 (Q2_K) | 613 MB | ~1.2 GB | Heavily compressed |
Vision
| Model | Download | Memory guide | Good for |
|---|---|---|---|
| SmolVLM 256M Instruct | 175 MB | ~600 MB | Fastest |
| SmolVLM 500M Instruct | 437 MB | ~1.0 GB | Balanced |
Voice
| Model | Download | Memory guide | Good for |
|---|---|---|---|
| Whisper Tiny (q8_0) | 44 MB | ~200 MB | FastestMultilingual |
| Whisper Tiny | 78 MB | ~250 MB | Multilingual |
| Whisper Base (q8_0) | 82 MB | ~250 MB | BalancedMultilingual |
| Whisper Base | 148 MB | ~350 MB | Multilingual |
| Whisper Small (q8_0) | 264 MB | ~500 MB | Best hereMultilingual |
| Whisper Medium (q8_0) | 823 MB | ~1.4 GB | Best hereMultilingualHigh-end devices |
Documents
| Model | Download | Memory guide | Good for |
|---|---|---|---|
| Nomic Embed Text v1.5 (Q4_0) | 77 MB | ~250 MB | Document Q&A |
Images
| Model | Download | Memory guide | Good for |
|---|---|---|---|
| DreamShaper 8 LCM (Q4_0) | 1.6 GB | ~3.4 GB | Fastest |
| Stable Diffusion v1.5 (Q4_0) | 1.8 GB | ~3.6 GB | Balanced |
| SD Turbo v2.1 (Q4_0) | 2.2 GB | ~4.3 GB | Fastest |
| DreamShaper XL v2 Turbo (Q4_K_S) | 2.8 GB | ~5.2 GB | Best hereHigh-end devices |
Video
| Model | Download | Memory guide | Good for |
|---|---|---|---|
| Wan 2.1 T2V 1.3B (Q4_0) | 866 MB | ~2.8 GB | Needs extra files |
| Wan 2.2 TI2V 5B (Q4_0) | 3.0 GB | ~6.0 GB | High-end devicesNeeds extra files |
| Wan 2.1 I2V 14B 480P (Q4_0) | 10.2 GB | ~16.8 GB | High-end devicesNeeds extra files |
Download sizes are the ones in the app's catalog, which can change over time. The memory figure is a rough guide to the free RAM a model needs while running; real use varies with your device and settings.
What to expect
Models that fit on a phone are small. They answer fast and privately, but they make mistakes, know less about the world than cloud models and cannot search the web. Use them for drafting, summarizing, translating and brainstorming, and double-check important facts. Bigger models answer better but need more memory and storage.
Bring your own model
The app can also import any GGUF chat model from your files or download one from a link, for example from Hugging Face. Larger models give better answers on laptops and desktops with plenty of RAM.