Cohere Command R+ (08-2024)
Overview
Model Name: Cohere: Command R+ (08-2024)
Model ID: cohere/command-r-plus-08-2024
Creator: Cohere
Release Date: August 30, 2024
Last Updated: November 10, 2025
An enhanced iteration of Command R+ delivering roughly 50% higher throughput and 25% lower latencies as compared to the previous Command R+ version while maintaining equivalent hardware requirements.
Specifications
| Specification | Value |
|---|---|
| Context Length | 128,000 tokens |
| Max Completion Tokens | 4,000 |
| Input Modalities | Text |
| Output Modalities | Text |
| Reasoning Support | No |
Pricing
| Metric | Cost |
|---|---|
| Input | $2.50 per 1M tokens ($0.0000025 per token) |
| Output | $10.00 per 1M tokens ($0.00001 per token) |
| Image Processing | $0 |
| Request Fee | $0 |
| Web Search | $0 |
Performance
Key Improvements
- 50% higher throughput compared to previous version
- 25% lower latencies compared to previous version
- Maintains equivalent hardware requirements
Usage Analytics (December 23, 2025)
- Daily requests: Ranges from 379 to 53,014 requests
- Recent sample: 5,112 requests
- Prompt tokens processed: 1,739,704
- Completion tokens processed: 118,225
Capabilities
- Structured outputs support
- Tool use and function calling
- Error tracking for tool calls
Related Models
Command R+ (Previous Version)
- Model ID: cohere/command-r-plus
- Predecessor to the current 08-2024 version
- 50% lower throughput and 25% higher latencies
Providers
Primary Provider: Cohere
- Base URL: https://api.langmart.ai/v1
- Adapter: CohereV2Adapter
- Status Page: https://status.cohere.ai/
Parameters
The model supports the following configuration parameters:
max_tokens- Maximum number of completion tokenstemperature- Sampling temperature for randomness controltop_p- Nucleus sampling parameterstop- Stop sequencesfrequency_penalty- Frequency penalty parameterpresence_penalty- Presence penalty parametertop_k- Top-k sampling parameterseed- Random seed for reproducibilityresponse_format- Response format specificationstructured_outputs- Structured output format supporttools- Tool/function definitionstool_choice- Tool selection strategy
Data Policy
| Policy | Status |
|---|---|
| Training on Prompts | Not permitted |
| Prompt Retention | 30 days |
| Publication Rights | Restricted |
| Usage Terms | Subject to Cohere's Usage Policy and SaaS Agreement |
Resources
- [Chat Interface](https://langmart.ai/chat
- [Model Comparison Tool](https://langmart.ai/model-docs
- Launch Announcement
- Cohere Usage Policy
- Cohere SaaS Agreement