What is Cohere?
Cohere builds enterprise AI: language models, embedding models and rerank models that are available through an API, along with the North platform. Its site positions the products as private, secure and customizable, for moving from proof of concept into production within existing systems.
Recent product announcements on the homepage include Embed 5, a family of embedding models, Rerank 5 with Context Focus for search, and North 2, a platform update. Model Vault provides fully isolated inference with the simplicity of a SaaS service.
API access works through two key types. Trial API keys are free but rate limited and not permitted for production or commercial use. Production keys are billed on a pay-as-you-go basis, per million tokens for language models; for example, Command R+ 08-2024 costs $2.50 per million input tokens and $10.00 per million output tokens on the pricing FAQ. A new account starts as a personal account with trial keys only, and production access is requested from the billing page of the dashboard.
Key capabilities
- Trial API keys Free, rate limited, not for production or commercial use.
- Production API keys Billed pay-as-you-go.
- Embed and Rerank Embedding models and a rerank model for search.
- Model Vault Fully isolated inference with SaaS simplicity.
- North Cohere's enterprise platform.
Where Cohere fits
- Command Generative language models for text generation, billed per million tokens on production API keys.
- Embed A semantic representation model that turns text into embeddings for search and retrieval, with Embed 5 announced as a new family.
- Rerank A retrieval optimization model that reorders search results; Rerank 5 adds Context Focus.
- Parse and Transcribe A document parsing model and a speech recognition model for feeding documents and audio into the same workflows.
- Model Vault A dedicated model inference platform with fully isolated inference and SaaS simplicity.
- North and Compass North is an agentic enterprise AI platform you control, and Compass handles intelligent search and discovery.
- Private deployments Run Cohere on-prem or in isolated VPCs when data cannot leave your own environment.
Reviews of Cohere
Be the first to share your experience with Cohere.
Write a review
Loading review form…
Alternatives to Cohere
AnyscaleAnyscale is the company behind Ray. Its managed platform runs training, batch inference and LLM serving on autoscaling Ray clusters, hosted or BYOC.
DeepInfraDeepInfra is an AI inference cloud that serves 200+ open and hosted models through OpenAI- and Anthropic-compatible APIs, billed per token or execution time, plus dedicated GPU deployments, GPU instances, sandboxes and batch jobs.
Fireworks AIFireworks AI runs open models through pay-per-token serverless inference, reserved throughput and inference inside your own cloud, plus training and routing.