Skip to content
SupportLogin
AutoScientist

Supported models

Base models available to AutoScientist and the per-model limits for context length, batch size, and LoRA rank.

Every model below can be passed as the model field when creating an AutoScientist run. The runtime model list is authoritative; use this page as a planning reference.

response = client.autoscientist.list_models()
for model in response.models:
print(model.id, model.api_model_size, model.methods, model.training_types)

See the autoscientist.list_models API reference for the current response schema.

Modelmodel IDSizeMethodsTraining types
Qwen 3.5 (0.8B)Qwen/Qwen3.5-0.8B0.8BSFT, DPOLoRA
Meta Llama 3.2 (3B Instruct)meta-llama/Llama-3.2-3B-Instruct3BSFT, DPOLoRA, full
Gemma 3 (4B Instruct)google/gemma-3-4b-it4BSFT, DPOLoRA, full
Gemma 3 (4B) VLMgoogle/gemma-3-4b-it-VLM4BSFT, DPOLoRA
Mistral (7B) Instruct v0.2mistralai/Mistral-7B-Instruct-v0.27BSFT, DPOLoRA
Qwen 3.5 (9B) †Qwen/Qwen3.5-9B9BSFT, DPOLoRA
gpt-oss (20B)openai/gpt-oss-20b20BSFT, DPOLoRA
Gemma 3 (27B) VLMgoogle/gemma-3-27b-it-VLM27BSFT, DPOLoRA
Gemma 4 (31B Instruct)google/gemma-4-31B-it31BSFT, DPOLoRA, full
Gemma 4 (31B) VLMgoogle/gemma-4-31B-it-VLM31BSFT, DPOLoRA
Mixtral 8x7B Instruct (v0.1)mistralai/Mixtral-8x7B-Instruct-v0.146.7BSFT, DPOLoRA, full
Llama 3.3 (70B Instruct)meta-llama/Llama-3.3-70B-Instruct-Reference70BSFT, DPOLoRA, full
Llama 4 Scout (17B-16E Instruct)meta-llama/Llama-4-Scout-17B-16E-Instruct109BSFT, DPOLoRA
gpt-oss (120B)openai/gpt-oss-120b120BSFT, DPOLoRA

VLM variants accept image columns as multimodal context. Qwen/Qwen3.5-9B (†) is also multimodal despite not carrying the suffix. A full training type means full fine-tuning is available in addition to LoRA.

These limits are properties of the model and training hardware. AutoScientist accounts for them automatically; they matter primarily when you override its recommended hyperparameters.

model IDContext (SFT)Context (DPO)Batch sizeGradient accumulationMaximum lora_r
Qwen/Qwen3.5-0.8B1310721310728164
Qwen/Qwen3.5-9B65536491528164
openai/gpt-oss-20b131072655361864
google/gemma-4-31B-it49152245764264
mistralai/Mixtral-8x7B-Instruct-v0.132768163848164
meta-llama/Llama-3.3-70B-Instruct-Reference24576122888164
meta-llama/Llama-4-Scout-17B-16E-Instruct65536122888164
openai/gpt-oss-120b65536327682864

Limits for google/gemma-3-4b-it, google/gemma-3-4b-it-VLM, google/gemma-3-27b-it-VLM, and meta-llama/Llama-3.2-3B-Instruct are not published. Leave batch_size unset so AutoScientist resolves the correct value.

Each model has one valid batch size. batch_size defaults to "max", which resolves to that model-specific value. Leaving it unchanged is the right choice for almost every run. Supplying another integer causes the launch to fail after the job is accepted.

Models use gradient accumulation when a larger effective batch is useful. This is why gpt-oss trains with a physical batch size of 1–2 and gradient accumulation of 8.

# Recommended: let AutoScientist resolve the model and batch size.
run = client.autoscientist.create(dataset_id=dataset_id)
# Advanced: pin a model and its required batch size.
run = client.autoscientist.create(
dataset_id=dataset_id,
model="openai/gpt-oss-20b",
hyperparams={"batch_size": 1},
)

lora_r accepts values from 1 through 64 for every model on this page. lora_alpha must be exactly one or two times lora_r; the API cross-checks the values when both are supplied in the same request.

AutoScientist requires at least 1,000 rows for every supported model. A smaller model does not lower this platform-wide minimum.

Training context is the maximum sequence length used during fine-tuning. It is shorter than the serving context and shorter for preference training than supervised fine-tuning.

Rows longer than the training limit are truncated rather than rejected. For datasets containing long documents, choose a model with a larger SFT or DPO context limit.