BasetenDecision modelOpen weights and API
Bison 1
Summary
Bison 1 is Baseten's decision model for structured Choice and Score outputs. The dense 31-billion-parameter checkpoint is based on Gemma 4 31B IT and accepts text or images.
On this page
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 7 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.
ClaimBench: Direct support (Show measurement, test conditions, and source)
- Source value
- 100
- Metric
- accuracy
- Unit
- %
- Benchmark version
- 1
- Category
- finance
- Task
- direct_support
- Task label
- Direct support
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Vals AI
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Source model ID: decision/bison-1. Source model name: Bison 1. Provider reported by source: Baseten. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-20. Maximum output tokens: 0. Latency in seconds: 0.866. Standard error: 0. Cost per task in USD: 0.000499. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-21-5dc3eba20db7. Source payload hash: 5dc3eba20db74e5214c71d077ff3785a5201c2ca4303917a1b122f03665d8562. Parser version: vals-astro-v4. Imported at: 2026-11-21. Source cohort: vals-claimbench:1:direct_support
ClaimBench: Missing evidence (Show measurement, test conditions, and source)
- Source value
- 100
- Metric
- accuracy
- Unit
- %
- Benchmark version
- 1
- Category
- finance
- Task
- missing_evidence
- Task label
- Missing evidence
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Vals AI
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Source model ID: decision/bison-1. Source model name: Bison 1. Provider reported by source: Baseten. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-20. Maximum output tokens: 0. Latency in seconds: 0.754. Standard error: 0. Cost per task in USD: 0.000527. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-21-5dc3eba20db7. Source payload hash: 5dc3eba20db74e5214c71d077ff3785a5201c2ca4303917a1b122f03665d8562. Parser version: vals-astro-v4. Imported at: 2026-11-21. Source cohort: vals-claimbench:1:missing_evidence
ClaimBench: Non-supporting citation (Show measurement, test conditions, and source)
- Source value
- 100
- Metric
- accuracy
- Unit
- %
- Benchmark version
- 1
- Category
- finance
- Task
- non_supporting_citation
- Task label
- Non-supporting citation
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Vals AI
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Source model ID: decision/bison-1. Source model name: Bison 1. Provider reported by source: Baseten. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-20. Maximum output tokens: 0. Latency in seconds: 0.695. Standard error: 0. Cost per task in USD: 0.000525. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-21-5dc3eba20db7. Source payload hash: 5dc3eba20db74e5214c71d077ff3785a5201c2ca4303917a1b122f03665d8562. Parser version: vals-astro-v4. Imported at: 2026-11-21. Source cohort: vals-claimbench:1:non_supporting_citation
ClaimBench: Numeric discrepancy (Show measurement, test conditions, and source)
- Source value
- 99.746
- Metric
- accuracy
- Unit
- %
- Benchmark version
- 1
- Category
- finance
- Task
- numeric_discrepancy
- Task label
- Numeric discrepancy
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Vals AI
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Source model ID: decision/bison-1. Source model name: Bison 1. Provider reported by source: Baseten. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-20. Maximum output tokens: 0. Latency in seconds: 0.787. Standard error: 0.254. Cost per task in USD: 0.000538. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-21-5dc3eba20db7. Source payload hash: 5dc3eba20db74e5214c71d077ff3785a5201c2ca4303917a1b122f03665d8562. Parser version: vals-astro-v4. Imported at: 2026-11-21. Source cohort: vals-claimbench:1:numeric_discrepancy
ClaimBench: Overall (Show measurement, test conditions, and source)
- Source value
- 99.082
- Metric
- accuracy
- Unit
- %
- Benchmark version
- 1
- Category
- finance
- Task
- overall
- Task label
- Overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Vals AI
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Source model ID: decision/bison-1. Source model name: Bison 1. Provider reported by source: Baseten. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-20. Maximum output tokens: 0. Latency in seconds: 0.799. Standard error: 0.216. Cost per task in USD: 0.000518. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-21-5dc3eba20db7. Source payload hash: 5dc3eba20db74e5214c71d077ff3785a5201c2ca4303917a1b122f03665d8562. Parser version: vals-astro-v4. Imported at: 2026-11-21. Source cohort: vals-claimbench:1:overall
ClaimBench: Subtle contradiction (Show measurement, test conditions, and source)
- Source value
- 95.663
- Metric
- accuracy
- Unit
- %
- Benchmark version
- 1
- Category
- finance
- Task
- subtle_contradiction
- Task label
- Subtle contradiction
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Vals AI
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Source model ID: decision/bison-1. Source model name: Bison 1. Provider reported by source: Baseten. Evaluator: Vals AI. Dataset type: public. Leaderboard date: 2026-11-20. Maximum output tokens: 0. Latency in seconds: 0.891. Standard error: 1.03. Cost per task in USD: 0.000501. Source ID: vals-ai. Evidence capture ID: vals-claimbench-2026-11-21-5dc3eba20db7. Source payload hash: 5dc3eba20db74e5214c71d077ff3785a5201c2ca4303917a1b122f03665d8562. Parser version: vals-astro-v4. Imported at: 2026-11-21. Source cohort: vals-claimbench:1:subtle_contradiction
Artificial Analysis Decision Index v2.1 (Show measurement, test conditions, and source)
- Source value
- 76
- Metric
- Decision Index, weighted probability-sensitive composite of six evaluations, 0-100 scale
- Unit
- points
- Benchmark version
- Artificial Analysis Decision Index v2.1
- Category
- decision
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Artificial Analysis
- Status
- active
- Retrieved at
- 2026-11-21
- Methodology
- Independent evaluation. Artificial Analysis lists the 31B checkpoint as Bison 31B. The index weights six evaluations and uses the predicted probabilities. This result applies to the deployment tested by Artificial Analysis.
Similar models
More models of the same kind as Bison 1, chosen by model line, developer, size, price and release date.
- Groundhog 1Baseten · same developer
- ClefCloudflare · similar size
- AutoTrust JEV-27B-VLAutoTrust AI · similar size
- AutoTrust JEV-27BAutoTrust AI · similar size
- pplx-decider-v1.1-27bPerplexity · similar size
- pplx-decider-v1.2Perplexity · similar size
- Drex v2.0Nace.AI · similar size
- Drex MaxNace.AI · similar size
All specifications
Architecture
- Model class
- A structured decision model for Choice and Score outputs, not a free-form text generator.
- Source: Model class
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
- License
- Baseten's model modifications are under MIT. Google's Apache 2.0 terms apply to the underlying Gemma weights.
- Source: License
- Shutdown
- None announced
- no source
API
Further details
Prices
- Baseten API price
- $0.10 per 1M tokensSource
Sources and data date
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| Vals AI | 1 | 0 | 6 |
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 21, 2026. 3 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Additional source | Baseten Bison 1 announcement (retrieved November 21, 2026) · Baseten: Bison 1 and Groundhog 1 · Editorial description reviewed November 21, 2026 |
| Additional source | Baseten Bison 1 and Groundhog 1 availability (retrieved November 21, 2026) |
| Additional source | Bison 1 model card (retrieved November 21, 2026) · Bison 1 model card · Editorial description reviewed November 21, 2026 |
| Additional source | Bison 1 license (retrieved November 21, 2026) |
| Additional source | Vals AI (retrieved November 21, 2026) |
| Additional source | Artificial Analysis (retrieved November 21, 2026) |
| Additional source | Model metadata (retrieved November 21, 2026) |