Supported models
The base models Ertas can fine-tune today, with licenses, parameter counts, GPU tier, GGUF size, and notes on what each is best for.
This page is the single-table view of every base model currently in the Ertas catalogue. For the full marketing-grade view (license excerpts, taxonomy, dataset compatibility notes) see the models index. For the size math behind the GGUF column, see File sizes and formats. For why a 5B total-parameter model can need A10G even when the compute-parameter count is lower, see Concepts on Gemma 4 E2B.
The catalogue
GGUF sizes are at Q4_K_M (the only quantisation Ertas exports today, see Quantization). "Free plan?" answers whether a Free-plan account can train this model; A10G models require Builder or higher.
Llama family
| Model | Params | License | GGUF size (Q4_K_M) | GPU tier | Free plan? | Best for |
|---|---|---|---|---|---|---|
| Llama 3.2 1B Instruct | 1.0B | Llama Community License | 0.81 GB | T4 | Yes | Smallest official Llama; web and mid-range mobile |
| Llama 3.2 3B Instruct | 3.0B | Llama Community License | 2.02 GB | T4 | Yes | Mobile and desktop sweet spot |
| Llama 3.1 8B Instruct | 8.0B | Llama Community License | 4.92 GB | A10G | No | Desktop and server-side inference where the extra quality matters |
Mistral family
| Model | Params | License | GGUF size (Q4_K_M) | GPU tier | Free plan? | Best for |
|---|---|---|---|---|---|---|
| Mistral 7B Instruct | 7.0B | Apache 2.0 | 4.37 GB | A10G | No | Permissively-licensed general purpose |
Mixtral and other mixture-of-experts variants are not yet trainable in Ertas; see Known limitations.
Phi family
| Model | Params | License | GGUF size (Q4_K_M) | GPU tier | Free plan? | Best for |
|---|---|---|---|---|---|---|
| Phi-3 mini 4k Instruct | 3.8B | MIT | 2.39 GB | T4 | Yes | Permissive license, small footprint, reasonable quality |
| Phi-4 Mini Instruct | 3.84B | MIT | ~2.4 GB | T4 | Yes | Newer Phi generation; stronger instruction-following than Phi-3 mini at the same class |
Gemma family
| Model | Params | License | GGUF size (Q4_K_M) | GPU tier | Free plan? | Best for |
|---|---|---|---|---|---|---|
| Gemma 3 1B IT | 1.0B | Gemma Terms of Use | 0.81 GB | T4 | Yes | Smallest Gemma; rough parity with Llama 3.2 1B |
| Gemma 3 4B IT | 4.0B | Gemma Terms of Use | 2.49 GB | T4 | Yes | Slightly larger than Llama 3.2 3B; same class |
| Gemma 4 E2B | 5.1B total / 2.3B effective | Gemma Terms of Use | 3.19 GB | A10G | No | Higher quality at 3B-class inference; needs A10G due to embedding tables |
Qwen family
| Model | Params | License | GGUF size (Q4_K_M) | GPU tier | Free plan? | Best for |
|---|---|---|---|---|---|---|
| Qwen 2.5 0.5B Instruct | 0.5B | Apache 2.0 | 0.49 GB | T4 | Yes | Tiny baseline; useful for testing pipelines, not production |
| Qwen 2.5 1.5B Instruct | 1.5B | Apache 2.0 | 1.12 GB | T4 | Yes | Multilingual support, permissive license |
| Qwen 2.5 3B Instruct | 3.0B | Apache 2.0 | 2.10 GB | T4 | Yes | Strong structured-output and multilingual at 3B class |
| Qwen 2.5 7B Instruct | 7.0B | Apache 2.0 | 4.68 GB | A10G | No | High-quality multilingual and long-context |
| Qwen 2.5 14B Instruct | 14.0B | Apache 2.0 | 8.99 GB | A10G | No | Largest catalogue model; desktop-only deployment |
Code-specialised
| Model | Params | License | GGUF size (Q4_K_M) | GPU tier | Free plan? | Best for |
|---|---|---|---|---|---|---|
| Qwen 2.5 Coder 1.5B | 1.5B | Apache 2.0 | 1.12 GB | T4 | Yes | Lightweight code completion; FIM support out of the box |
| Qwen 2.5 Coder 3B | 3.0B | Apache 2.0 | 2.10 GB | T4 | Yes | Code completion recipe default; see Cookbook: code completion |
| Qwen 2.5 Coder 7B | 7.0B | Apache 2.0 | 4.68 GB | A10G | No | Desktop coding assistant tier |
Under active evaluation (not yet in catalogue)
Models in this group are being tested and validated for catalogue inclusion. They are not yet selectable in the Studio model picker; the models index is the source of truth for what is currently selectable.
| Model | Params | License | Status |
|---|---|---|---|
| Llama 4 (when generally available) | TBD | Llama Community License | Under evaluation |
| Phi-4 (full, ~14B) | ~14B | MIT | Under evaluation |
| Qwen 3 family | TBD | Apache 2.0 | Under evaluation |
| TinyLlama 1.1B Chat | 1.1B | Apache 2.0 | Under evaluation; useful as a very-low-footprint sub-3B candidate |
| SmolLM 1.7B Instruct | 1.7B | Apache 2.0 | Under evaluation; sub-3B with stronger instruction following than TinyLlama |
If you are relying on one of these for an in-flight project, the Hugging Face URL import path lets you train against them as unverified architectures today, at the cost of giving up automatic credit refunds on training failures. Catalogue inclusion will lift the unverified status when validation completes.
Picking from the catalogue
If you are deciding between models, the right path is Picking a base model. Three quick heuristics that cover most cases:
- For mobile, default to a 3B-class model at Q4_K_M (Llama 3.2 3B, Qwen 2.5 3B, Phi-3 mini, Gemma 3 4B). Below 1B is rough for instruction following; above 3B starts to crowd phone RAM.
- For desktop, an 8B-class model (Llama 3.1 8B, Mistral 7B, Qwen 2.5 7B) is a noticeable quality bump over 3B and still fits in 5 GB on disk.
- For web, prefer 1B-class (Llama 3.2 1B, Gemma 3 1B, Qwen 2.5 1.5B) to keep first-load downloads under 1 GB and stay inside browser memory ceilings.
Hugging Face models outside the catalogue
You can fine-tune any Unsloth-compatible model from a Hugging Face URL. Ertas validates the architecture and reports whether the model is known to fine-tune cleanly. If validation is uncertain, the run still queues but credits are not refunded on training failures for unverified architectures. See Picking a base model for the bring-your-own-model path.
License notes
Every catalogue model's license is summarised here, but you must read the full license before shipping a commercial product. Notable distinctions:
- Apache 2.0 (Qwen, Qwen 2.5 Coder, Mistral 7B): permissive, commercial use allowed without extra conditions.
- MIT (Phi-3 mini, Phi-4 Mini): permissive, commercial use allowed.
- Llama Community License (Llama 3.1, Llama 3.2): commercial use allowed up to a monthly active user threshold; products over the threshold need a separate Meta agreement. The threshold is documented in Meta's license text and changes occasionally.
- Gemma Terms of Use (Gemma 3, Gemma 4): commercial use allowed with specific prohibited-use restrictions Google maintains separately. Read both the Terms of Use and the Prohibited Use Policy.
Ertas does not enforce license compliance; the model picker lists the license on each card and the models index carries the full text excerpt per model.