A
AI21: Jamba Large 1.7
● Excellent8.2
/ 10
Jamba Large 1.7 is the latest model in the Jamba open family, offering improvements in grounding, instruction-following, and overall efficiency. Built on a hybrid SSM-Transformer architecture with a 256K context...
Specifications
| Attribute | Value |
|---|---|
| Lab | Ai21 |
| Tags | Intelligent Long Context |
| Overall Score | 8.2/10 |
| Release Date | 2025-08 |
| Context Window | 256,000 tokens |
| Input Price / 1M | $2.00 |
| Output Price / 1M | $8.00 |
| Input Modalities | Text |
| Output Modalities | Text |
Strengths
- Hybrid SSM-Transformer architecture for efficient long-context processing
- 256K context window with strong grounding and instruction following
- Frontier-lite pricing ($2.00/$8.00) with flagship-competitive quality
- Improved efficiency over previous Jamba versions
- Strong at multi-document reasoning and retrieval tasks
Weaknesses
- Higher pricing than comparable open-weight alternatives
- Less widely adopted than Transformer-only architectures
- Smaller community and fewer fine-tuned variants available
Best For
- Long-document analysis and summarization
- Enterprise RAG and knowledge retrieval
- Grounded instruction-following tasks
- Cost-effective alternative to flagship models
Sources & Further Reading
Related Models
Scores are aggregated from public benchmarks (MMLU, HumanEval, MATH, GSM8K, LMSYS) and normalized to a 1–10 scale. Methodology →