LLM Hallucination Index RAG SpecialModel Insights
command-r-plus
Details
Developer
Cohere
License
NA (private model)
Model parameters
104b
Supported context length
128k
Price for prompt token
$3/Million tokens
Price for response token
$15/Million tokens
Model Performance Across Task-Types
Chainpoll Score
Short Context
0.86
Medium Context
1
Long Context
0.95
Model Insights Across Task-Types
Digging deeper, here’s a look how command-r-plus performed across specific datasets
Short Context RAG
Medium Context RAG
This heatmap indicates the model's success in recalling information at different locations in the context. Green signifies success, while red indicates failure.
Long Context RAG
This heatmap indicates the model's success in recalling information at different locations in the context. Green signifies success, while red indicates failure.
Performance Summary
| Tasks | Task insight | Cost insight | Dataset | Context adherence | Avg response length |
|---|---|---|---|---|---|
| Short context RAG | The model demonstrates good reasoning and comprehension skills, excelling at short context RAG. It also decent mathematical proficiency, as evidenced by its performance on DROP and ConvFinQA benchmarks. It could have performed better for its size. | It is costly model for the performance it offers. We recomend using 30x cheaper Gemini Flash or 20x cheaper Haiku. | Drop | 0.81 | 238 |
| Hotpot | 0.88 | 238 | |||
| MS Marco | 0.89 | 238 | |||
| ConvFinQA | 0.88 | 238 | |||
| Medium context RAG | Flawless performance making it suitable for any context length upto 25000 tokens. | Great performance but we recommed using 30x cheaper Gemini Flash. | Medium context RAG | 1.00 | 238 |
| Long context RAG | Great performance with minor performance issues for all context length. | Good performance but we recommed using similar price Claude 3.5 Sonnet for best performance or 9x cheaper Gemini Flash for cost effective performance. | Long context RAG | 0.95 | 238 |