All benchmarks
AI21 Labs · Israel · Released 2024-08

Jamba 1.5 Large

Hybrid SSM-Transformer (Mamba + attention) with a very long 256K context and open weights.

Access
open-weights
Context
256,000 tokens
Modalities
text
Languages
9

Benchmark scores

MMLU
81.2%
Massive Multitask Language Understanding
GPQA
N/A
Graduate-level science QA (Diamond)
HumanEval
71.4%
Python code completion
SWE-bench
N/A
Real GitHub issue resolution (Verified)
LiveCodeBench
N/A
Contamination-free coding
AIME
N/A
American Invitational Mathematics Exam
MMMU
N/A
Multimodal college-level reasoning
MATH
N/A
Competition-level math word problems

Scores as reported by the vendor or leading public leaderboards. "N/A" means the score has not been publicly disclosed for this metric.

Explain like I'm 5

A hybrid model that reads super long documents fast.

Key features
  • SSM + Transformer hybrid
  • 256K context
Strengths
  • Long context efficiency
Limitations
  • Below frontier on reasoning

Best for

Long-doc RAGSummarization

Related models