DeepSeek V4.1 Flash
Catalog · 552B + 196BBackbone + Engram memory · 8B active prefill / 16B decode · Vision, text and agents
Farka separates license review, hardware fit, live training, paired evaluation, tuned inference, export, billing, and teardown. A model is not called verified because it appeared in a list or accepted a quote.
Open models for reasoning, code, and agents. The training setup, capacity, and price depend on your workload. These catalog entries carry no completed-training or quality claim.
Backbone + Engram memory · 8B active prefill / 16B decode · Vision, text and agents
13B active parameters · Coding and agents
Reasoning, code and agents
Images, text and assistants
Exact model, training method, passes over the data, accelerator-memory class, and delivery combinations from collected runs. Farka uses different training approaches depending on the run: a light tune adds a small set of new weights to the base model, while a full tune updates every weight. Provider-native hardware remains attached to the audited run record.
Light tune · 3 passes
Light tune · 1 pass
Full tune · 3 passes
Light tune · 3 passes
Light tune · 3 passes
Results are configuration-specific. They do not imply that every dataset, model size, or training recipe will improve.
Curated, license-reviewed models verified in production. Farka selects the cheapest verified capacity with sufficient memory.
Models outside the verified set, multi-accelerator, private-network, or customer-cloud workflows begin with a capacity and security review. Explore custom training.
Commercial use and weight modification must be supported by publisher evidence. Unknown, non-commercial, or incompatible licenses fail closed.