The Orchestrator › Models › Baseten
Baseten: Mammoth-1
Mammoth-1 is a 27B decision model from Baseten's Base Labs, fine-tuned from Qwen3.5-27B for classification, routing, and scoring. It answers multiple-choice, yes/no, and ordered scoring questions with probabilities over predefined options rather than generated text. Shared-prefix caching reuses context across questions, while per-option probabilities support confidence thresholds and escalation decisions.
Specifications
| Provider | Baseten |
| Model ID | baseten/mammoth-1 |
| Context window | 16K tokens (16,384) |
| Modalities | text → text |
| Tokenizer | — |
| Router sources | Vercel AI Gateway |
| Released | 2026-10-21 |
| Tracked since | 2026-11-14 |
| Popularity | Not in the ranked set |
| Reasoning | — |
Pricing (per million tokens)
| Variant | Source | API ID | Input | Output | Context |
|---|---|---|---|---|---|
| standard | Vercel AI Gateway | baseten/mammoth-1 | $0.1 | Free | 16K |
Recorded price history
| Date | Input / 1M | Output / 1M | Context |
|---|---|---|---|
| 2026-11-14 | $0.1 | Free | 16K |
Capabilities
Text in, text out — no special capability markers in the catalog data
Trust flags come from TrustedRouter's per-provider attestations.
Supported parameters
—