Models
All models are optimized with our JOBIM J-FACTOR compression technology, delivering state-of-the-art performance at a fraction of the cost.
JOBIM Utility
jobim-jfactor-utilityOur most efficient model for general tasks. Perfect for chatbots, content generation, and data processing.
Context
128K tokens
Cost
$0.10 / 1M tokens
Compression
98.2%
Throughput
13.6 TPS
JOBIM Llama 70B
jobim-jfactor-llama-70bHigh-performance model for complex reasoning, coding, and advanced tasks.
Context
131K tokens
Cost
$0.50 / 1M tokens
Compression
95.8%
Throughput
8.2 TPS
JOBIM Mistral 8x22B
jobim-jfactor-mistral-8x22bSpecialized model for reasoning, math, and scientific tasks with MoE architecture.
Context
64K tokens
Cost
$0.75 / 1M tokens
Compression
94.3%
Throughput
6.1 TPS
Model Comparison
| Model | Best For | Cost Efficiency | Performance |
|---|---|---|---|
| JOBIM Utility | General tasks, chatbots | ★★★★★ | ★★★★☆ |
| JOBIM Llama 70B | Complex reasoning, coding | ★★★★☆ | ★★★★★ |
| JOBIM Mistral 8x22B | Math, science, reasoning | ★★★☆☆ | ★★★★★ |