Cohere

New Embed 5 Pro model tier added to Model Vault: Small $3.00/hr ($2,000/mo) and Medium $5.00/hr ($3,250/mo) Model Vault Embed 4 Small replaced by Embed 5 Fast Small; hourly rate $4.00 to $3.00 and monthly $2,500 to $2,000 per instance; Embed 5 Fast Medium unchanged at $5.00/$3,250

New Parse 5 document-parsing model added to Model Vault, with Medium ($4.00/hr, $2,500/mo) and XL ($7.00/hr, $4,300/mo) performance tiers alongside existing Embed and Rerank models

Model Vault added a new 'Large' performance tier for Rerank 4 Pro at $10.00/hour ($6,500/month), alongside the existing Medium tier ($5.00/hour, $3,250/month)

A new 'Coding models' product category tab was added to the pricing page, expanding Cohere's offering from 3 to 4 product categories. Coding models joins Workplace systems, Generative models, and Advanced retrieval models.

Rerank 4 Pro Large performance tier added to Model Vault pricing table at $10.00/hr ($6,500/mo), expanding the available instance options.

Rerank 4 Pro Large tier removed from Model Vault pricing table. The $10.00/hr ($6,500/month) large performance tier is no longer listed.

Model Vault added as a new pricing tier with per-instance hourly and monthly rates. Embed 4 Small: $4.00/hr, $2,500/mo; Embed 4 Medium and all Rerank models: $5.00/hr, $3,250/mo.

Cohere restructured pricing page from transparent API pricing to enterprise workplace systems model - main page now emphasizes North and Compass with custom pricing instead of per-token costs

Embed 3 renamed to Embed 4 Embed text token pricing increased from $0.10 to $0.12 per 1M tokens (20% increase)

Command R7B replaced by Command A as the featured 'new' model - positioning changed from 'smallest generative model optimized for speed and efficiency' to 'most efficient and performant model to date, specializing in agentic AI, multilingual, and human evaluations'

Command R Fine-tuned Model output price decreased from $4.00 to $1.20 per 1M tokens (70% decrease) Command R output price decreased from $1.50 to $0.60 per 1M tokens (60% decrease)

Rerank 3 model introduced as new retrieval model at $2.00/1K searches Production tier plan removed from pricing page (previously offered custom model training and elevated support)

Command R input pricing decreased from $1.00/1M tokens to $0.50/1M tokens (50% reduction) Command R output pricing decreased from $2.00/1M tokens to $1.50/1M tokens (25% reduction)