Cohere
Custom Solution Available, Free Trial, Paid
Cohere’s top-tier large language models (LLMs), using retrieval-augmented generation (RAG), empower businesses to develop robust and secure applications capable of searching, comprehending context, and engaging in textual conversations with excellence.
Key Features
-
Research: Use semantic search capabilities for accurate and qualitative output
-
AI chat interface: Generate, summarize and classify texts
-
Customization: Develop and build your own AI applications and assistants that search, understand, and converse in text with precision
Pricing Structure
Free version available: Yes
Free, rate-limited usage for learning and prototyping, until going into production
Production: Train custom models, Elevated ticket support, Access to all endpoints, Increased rate limit
“Generate” and “Chat”:
Command: Input $1.00/1M tokens, Output $2.00/1M tokens
Command Light: Input $1.00/1M tokens, Output $2.00/1M tokens
Command Light (Fine-tuned model): Training $1.00/1M tokens, Input §0.30/1M tokens, Output $0.60/1M tokens
“Embed”: $0.10/1M tokens
“Classify”: Default and Fine-tuned model: $0.05/1K classifications
“Summarize”:
Command: Input $1.00/1M tokens, Output $2.00/1M tokens
“Rerank”:
Default and Fine-tuned model: $1.00/1K searchers
Enterprise: Custom price
Dedicated model instances, dedicated support channels, custom deployment options
Use Performance & Case Suitability
-
Cohere’s models use inference engines that deliver better runtime performance at a lower cost than open-source equivalents
-
For companies and developers looking to build their own text analysis applications
-
Supports 100+ languages
Customization & Flexibility
-
Cohere models have flexible deployment options. They are accessible through a SaaS API, cloud services (e.g., OCI, AWS SageMaker, Bedrock), and private deployments (VPC and on-prem)
-
Customers have complete control over customization and model inputs/outputs via fine-tuning tools
Data Handling & Privacy
- Customer data is not used in training base model
- Cloud agnostic, SOC 2 compliant, and focus on security, privacy, and Responsible AI
Support & Documentation
-
Contact via online forms and email
-
Dedicated support for enterprise customers
-
Discord community
-
LLM University