Jamba 1.5 Large

Zero-eval
#3Wild Bench

by AI21 Labs

+
+
+
+
About

Jamba 1.5 Large was developed by AI21 Labs using a hybrid architecture combining transformer and state space models, designed to provide efficient long-context understanding. Built to handle extended documents and conversations with computational efficiency, it represents AI21's innovation in efficient large-scale model design.

+
+
+
+
Pricing Range
Input (per 1M)$2.00 -$2.00
Output (per 1M)$8.00 -$8.00
Providers2
+
+
+
+
Timeline
AnnouncedAug 22, 2024
ReleasedAug 22, 2024
Knowledge CutoffMar 5, 2024
+
+
+
+
License & Family
License
Jamba Open Model License
Performance Overview
Performance metrics and category breakdown

Overall Performance

8 benchmarks
Average Score
65.5%
Best Score
93.0%
High Performers (80%+)
3

Performance Metrics

Max Context Window
512.0K
Avg Throughput
71.0 tok/s
Avg Latency
0ms
+
+
+
+
All Benchmark Results for Jamba 1.5 Large
Complete list of benchmark scores with detailed information
ARC-C
text
0.93
93.0%
Self-reported
GSM8k
text
0.87
87.0%
Self-reported
MMLU
text
0.81
81.2%
Self-reported
Arena Hard
text
0.65
65.4%
Self-reported
TruthfulQA
text
0.58
58.3%
Self-reported
MMLU-Pro
text
0.54
53.5%
Self-reported
Wild Bench
text
0.48
48.5%
Self-reported
GPQA
text
0.37
36.9%
Self-reported
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+