Cohere offers large language model APIs that allow businesses to integrate natural language processing into their applications seamlessly.
Pricing Change History
17 changesModel Vault Rerank 4 Fast/Pro Medium renamed Rerank 5 and cut from $5.00/hr ($3,250/mo) to $4.00/hr ($2,500/mo); Rerank 4 Pro Large ($10.00) and Rerank 3.5 rows removed.
Model Vault Embed 4 replaced by Embed 5 Fast/Pro; Small tier drops $4.00 to $3.00/hr ($2,500 to $2,000/mo); Medium unchanged at $5.00/$3,250; Embed 5 Pro Small/Medium added.
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
Weekly roll-up · includes Aug 28 daily
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)
Quarterly roll-up · includes Jun 15 weekly, May 4 weekly
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.
Quarterly roll-up · includes Mar 30 weekly
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 text token pricing increased from $0.10 to $0.12 per 1M tokens (20% increase) Embed image pricing metric changed from '$0.0001 per 1 Image' to '$0.47 per 1M Image Tokens'
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'
Rerank document length threshold decreased from 4096 tokens to 500 tokens (documents longer than this limit are split into chunks) Rerank upgraded from version 3 to version 3.5
Embed 3 now supports multimodal image embeddings at $0.0001 per image Command R Fine-tuned Model training price decreased from $8.00 to $3.00 per 1M tokens (62.5% decrease)
Free tier plan removed from pricing page (previously offered rate-limited usage for learning and prototyping) Command R Fine-tuned Model pricing explicitly disclosed: $2.00/1M input tokens, $4.00/1M output tokens, $8.00/1M training tokens
Research Cohere's pricing with the Pulse Agent
Ask anything about Cohere's plans, packaging shifts, and how it compares across the market — answered live from verified data.
Ask the agentCurrent plans
Captured| Plan | Type | Price |
|---|---|---|
| Command A+ | Plan | Free |
| North Mini Code | Plan | Free |
| Embed 5 Fast | Plan | $0.08 Input tokens |
| Embed 5 Pro | Plan | $0.12 Input tokens |
| Parse 5 | Plan | $1.5 Pages |
| Rerank 4 Fast | Plan | $2 Searches |
| Rerank 4 Pro | Plan | $2.5 Searches |
| Command R7B | Plan | $0.15 Output tokens + $0.04 Input tokens |
| Compass | Plan | Contact |
| Command R | Plan | $0.15 Input tokens + $0.6 Output tokens |
| North | Plan | Contact |
| Transcribe | Plan | $3.75 per instance |
| Model Vault - Parse 5 XL | Add-on | $4,300 /mo + usage |
| Model Vault - Embed 5 Pro Medium | Add-on | $3,250 /mo + usage |
| Model Vault - Rerank 4 Fast Medium | Add-on | $3,250 /mo + usage |
| Model Vault - Embed 5 Fast Medium | Add-on | $3,250 /mo + usage |
| Model Vault - Embed 5 Pro Small | Add-on | $2,000 /mo + usage |
| Model Vault - Rerank 4 Pro Large | Add-on | $6,500 /mo + usage |
| Model Vault - Rerank 3.5 Medium | Add-on | $3,250 /mo + usage |
| Model Vault - Parse 5 Medium | Add-on | $2,500 /mo + usage |
| Model Vault - Embed 5 Fast Small | Add-on | $2,000 /mo + usage |
| Model Vault - Rerank 4 Pro Medium | Add-on | $3,250 /mo + usage |
Company Facts
Pulse news about Cohere
View all →How Cohere compares with its peers
| Company | Entry price | Last change |
|---|---|---|
O OpenAI | $8 /mo | Oct 30, 2026 |
A Anthropic | $17 /mo, billed annually | Oct 12, 2026 |
G Google | Contact | Oct 12, 2026 |
M Metabase | $90 /mo, billed annually | Oct 26, 2026 |
M Mistral AI | $14.99 /mo | Oct 12, 2026 |
T Together AI | $9 per GPU-hour (B200, promo) | Oct 26, 2026 |
Cohere pricing FAQ
- How much does Cohere cost?
- Cohere pricing starts at $0.08 Input tokens on Embed 5 Fast. Cohere lists 12 plans on its pricing page; the Command A+ plan is free.
- Does Cohere have a free plan?
- Yes. Cohere's Command A+ plan is free; paid plans start at $0.08 Input tokens on Embed 5 Fast.
- What pricing model does Cohere use?
- Cohere uses subscription + usage pricing across 12 plans, with a free tier.
- When did Cohere last change its pricing?
- Oct 26, 2026: Cohere: Model Vault Rerank 4 Fast/Pro Medium renamed Rerank 5 and cut from $5.00/hr ($3,250/mo) to $4.00/hr….
- How often has Cohere changed its pricing?
- Pulse has recorded 17 pricing changes for Cohere since it started tracking on Dec 1, 2024; the pricing page was last captured on Oct 26, 2026.
- Who are Cohere's pricing peers?
- Pulse compares Cohere with OpenAI, Anthropic, Google, Metabase, Mistral AI and Together AI.