Language modelOpen source
Jamba2-Mini
AI21
- Released
- January 8, 2026
- Data date
- October 3, 2026
Jamba2-Mini uses mixture of experts within AI21’s hybrid SSM-Transformer architecture. It activates 12 of its 52 billion parameters per token. It launched alongside Jamba2-3B, also targeting answers that stay grounded in supplied sources.
AI21 specifies a 256K-token context window. The weights use Apache 2.0 and are offered through AI21 Studio and Hugging Face.
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.
IFBench (Show measurement, test conditions, and source)
- Source value
- 0.38
- Score
- 0.38
- Metric
- IFBench
- Unit
- ratio
- Category
- source-specific
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- AI21 announcement/HF chart
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Same four normalized benchmark components and Enterprise Reliability composite; separate 100-prompt blind human comparison.
- Context
- Source-specific observation; it is not a shared comparison cohort. AI21 self-reported comparison, not an external evaluator record
Collie (Show measurement, test conditions, and source)
- Source value
- 0.33
- Score
- 0.33
- Metric
- Collie
- Unit
- ratio
- Category
- source-specific
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- AI21 announcement/HF chart
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Same four normalized benchmark components and Enterprise Reliability composite; separate 100-prompt blind human comparison.
- Context
- Source-specific observation; it is not a shared comparison cohort. AI21 self-reported comparison, not an external evaluator record
IFEval (Show measurement, test conditions, and source)
- Source value
- 0.97
- Score
- 0.97
- Metric
- IFEval
- Unit
- ratio
- Category
- source-specific
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- AI21 announcement/HF chart
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Same four normalized benchmark components and Enterprise Reliability composite; separate 100-prompt blind human comparison.
- Context
- Source-specific observation; it is not a shared comparison cohort. AI21 self-reported comparison, not an external evaluator record
FACTS (Show measurement, test conditions, and source)
- Source value
- 0.57
- Score
- 0.57
- Metric
- FACTS
- Unit
- ratio
- Category
- source-specific
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- AI21 announcement/HF chart
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Same four normalized benchmark components and Enterprise Reliability composite; separate 100-prompt blind human comparison.
- Context
- Source-specific observation; it is not a shared comparison cohort. AI21 self-reported comparison, not an external evaluator record
Enterprise Reliability Score (average) (Show measurement, test conditions, and source)
- Source value
- 0.56
- Score
- 0.56
- Metric
- Enterprise Reliability Score (average)
- Unit
- ratio
- Category
- source-specific
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- AI21 announcement/HF chart
- Status
- active
- Retrieved at
- 2026-10-04
- Methodology
- Vendor-reported result. Same four normalized benchmark components and Enterprise Reliability composite; separate 100-prompt blind human comparison.
- Context
- Source-specific observation; it is not a shared comparison cohort. AI21 self-reported comparison, not an external evaluator record
Sources and data date
Every statement links to its underlying documentation or leaderboard.
| Type | Evidence and data date |
|---|---|
| Research status | Research date October 4, 2026. 1 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Additional source | AI21 Jamba2 announcement (retrieved October 3, 2026; October 4, 2026) · AI21 Jamba2 announcement · Editorial description reviewed October 4, 2026 |
| Additional source | AI21 Jamba2-Mini model card (retrieved October 3, 2026) · AI21 Jamba2-Mini model card · Editorial description reviewed October 4, 2026 |