Nace.AIDecision modelOpen weights
Drex v2.0
Summary
Drex v2.0 scores several questions against a shared text context. Given a document and a set of answers, it can help you classify an incoming request.
It uses the text backbone of Qwen3.6-35B-A3B, with around three billion of 35 billion parameters active per token. Local inference can use up to 131,072 context tokens. The API specifies a separate limit for the context and question together.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 2 published benchmark results
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
Decision Index 0.4.1 (Public) (Show measurement, test conditions, and source)
- Source value
- 61.25
- Score
- 61.25
- Metric
- chance-corrected skill
- Unit
- points
- Benchmark version
- 0.4.1
- Category
- decision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Nace.AI Drex technical report
- Status
- active
- Retrieved at
- 2026-11-07
- Methodology
- Vendor-reported result. Public-suite result from the October 22, 2026 leaderboard, reproduced in Table 12.
Decision Index 0.4.1 (Full) (Show measurement, test conditions, and source)
- Source value
- 52.76
- Score
- 52.76
- Metric
- chance-corrected skill
- Unit
- points
- Benchmark version
- 0.4.1
- Category
- decision
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Nace.AI Drex technical report
- Status
- active
- Retrieved at
- 2026-11-07
- Methodology
- Vendor-reported result. Full result including public and private tests from the October 22, 2026 leaderboard, reproduced in Table 12.
Similar models
More models of the same kind as Drex v2.0, chosen by model line, developer, size, price and release date.
- Drex MaxNace.AI · same developer
- Drex v1.5Nace.AI · same developer
- Drex DLMNace.AI · same developer
- AutoTrust JEV-27B-VLAutoTrust AI · similar size
- ClefCloudflare · similar size
- AutoTrust JEV-27BAutoTrust AI · similar size
- pplx-decider-v1.1-27bPerplexity · similar size
- Drex Max v1.1Nace.AI · same developer
All specifications
Architecture
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
- Shutdown
- None announced
- no source
API
Further details
- Local context window
- 16,384 tokens by default, configurable up to 131,072 tokens.
- Source: Local context window
- API input limit
- Up to 131,072 tokens of shared context and 262,144 tokens per question including context.
- Source: API input limit
Sources and data date
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Research status | Research date November 7, 2026. 1 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Additional source | Nace.AI API model catalog (retrieved November 7, 2026) · Nace.AI API model catalog · Editorial description reviewed November 7, 2026 |
| Additional source | Drex v2.0 model card (retrieved November 7, 2026) · Drex v2.0 model card · Editorial description reviewed November 7, 2026 |
| Additional source | Nace.AI Drex technical report (retrieved November 7, 2026) |