Cohere API Pricing (November 2026)
Cohere API prices run from $0.0375 to $1.50 per million tokens across 3 current models. The cheapest is Command R7B at $0.0375 input and $0.15 output per million tokens. The flagship, Command R, costs $0.15 input and $0.60 output per million tokens. Every rate below was read on Cohere's own pricing page, on the date shown in its row.
| Model | Input /M | Cache read /M | Output /M | Batch in /M | Batch out /M | Context | Status | Checked |
|---|---|---|---|---|---|---|---|---|
Command R7B command-r7b-12-2024 | $0.0375 | — | $0.15 | — | — | 128K | Current | 2026-10-28 |
Aya Expanse 32B c4ai-aya-expanse-32b | $0.50 | — | $1.50 | — | — | 128K | Current | 2026-10-28 |
Command R command-r-08-2024 | $0.15 | — | $0.60 | — | — | 128K | Flagship | 2026-10-28 |
Command R+ command-r-plus-08-2024 | $2.50 | — | $10.00 | — | — | 128K | Legacy | 2026-10-28 |
US dollars per million tokens, list rates on Cohere's own API. Cache read is the price of a cached input token. Batch is the asynchronous API, where the vendor offers one. The date in each row is when we last read it on the vendor's pricing page; it links to that page.
Cohere API monthly cost estimates
What light use (10M input + 2M output tokens a month), medium use (100M input + 20M output tokens a month), heavy use (1B input + 200M output tokens a month) costs on each current model, at list price with no caching and no batch discount.
| Model | Input / output per 1M | Light · 10M + 2M | Medium · 100M + 20M | Heavy · 1B + 200M |
|---|---|---|---|---|
| Command R7B | $0.0375 / $0.15 | $0.675 | $6.75 | $67.50 |
| Command R | $0.15 / $0.60 | $2.70 | $27.00 | $270 |
| Aya Expanse 32B | $0.50 / $1.50 | $8.00 | $80.00 | $800 |
Prompt caching and the Batch API cut these figures; the next sections show by how much. For your own token counts, use the LLM API cost calculator.
Cohere API price history
Cohere API price changes and releases
Model launches and pricing changes, dated from Cohere's blog, newsroom, developer changelog and pricing page.
Chat V2 with tool use and retrieval-augmented generation built in. Trial keys are free; production keys bill pay-as-you-go, invoiced at the end of each month or when the balance reaches $250.
DEDICATEDCohere's isolated dedicated inference, priced per instance by the hour or the month rather than per token. The public table covers Embed, Rerank and Parse instances; Command A+ and North Mini Code production deployments go through it too.
PRIVATEOn-premises or isolated VPC deployments for regulated industries, quoted by sales. No list price.
OPEN WEIGHTSCohereLabs publishes North Mini Code under Apache-2.0 alongside Cohere's other open-weight releases.
PRICINGPer-token cards for Command R and R7B, the Aya Expanse rate and legacy Command pricing in the FAQ, and the Model Vault instance table.
Mistral AI
The other enterprise lab selling sovereign and on-premises deployment, with a far larger public per-token catalog and open-weight models across sizes.
ENTERPRISEAnthropic
Claude is the usual alternative for document-heavy and agentic enterprise work; every current model has a public per-token rate, which Cohere's newest do not.
FRONTIER LABOpenAI
The reference for general-purpose API pricing, with small models priced in the same range as Command R7B.
OPEN-WEIGHT GIANTAlibaba (Qwen)
The Qwen (Tongyi Qianwen) model family from Alibaba Cloud's Tongyi Lab. First released in 2023, Qwen is the most-downloaded open-weight model family in the world — most tiers ship under Apache 2.0 on Hugging Face and ModelScope, while the proprietary Max tier is served through Alibaba Cloud's Model Studio.
GLM ARCHITECTZhipu (Z.ai / GLM)
The Beijing lab behind the GLM models, also operating internationally as Z.ai. Spun out of Tsinghua University in 2019, Zhipu ships strong coding/agentic models, multiple free tiers, and open weights — and was the first of China's "Six Tigers" to pursue an IPO. It also sits on the US Entity List.
MEDICAL-AI SPECIALISTBaichuan
The Beijing lab behind the Baichuan general models and the medical M-series. Founded in 2023 by Wang Xiaochuan (former CEO of search engine Sogou), Baichuan has pivoted hard toward healthcare AI — its open-source Baichuan-M3 tops several medical benchmarks — while keeping a general-purpose 4-series lineup.
Cohere API pricing FAQ
What Cohere's API costs today, which models have no token price, and what this hub leaves out.
Q · 01 How much does the Cohere API cost?
+
$0.15/M input and $0.60/M output, Command R7B at $0.0375/M and $0.15/M, and Aya Expanse at $0.50/M and $1.50/M, read October 29, 2026. No cache-read or batch rates are published. Q · 02 Why is Command A+ not priced here?
+
Q · 03 Is North Mini Code free?
+
Q · 04 What happened to Command R+?
+
Q · 05 Are trial API calls really free?
+
Q · 06 Are Embed, Rerank and Transcribe included?
+
Cohere was founded in Toronto in 2019 and has stayed an enterprise company: its products are the North AI platform, retrieval models (Embed, Rerank, Parse) and the Command family of text models, sold with private and dedicated deployment for regulated industries.
Public per-token pricing covers the older part of that catalog. Command R lists at $0.15 input and $0.60 output per 1M tokens, Command R7B at $0.0375 and $0.15, and Aya Expanse 32B at $0.50 and $1.50; none has a cache rate. Command R+ keeps its rate only for existing customers.
The newer models are priced differently. Command A+ (May 2026), the last of the Command A line, and North Mini Code (June 2026), Cohere's first open-source coding model, have no per-token rate on the pricing page: both are listed as free through an API key and as a download, under Apache-2.0. Cohere points production use of North Mini Code at Model Vault, which bills per instance. That is why this hub's lineup stops at the Command R generation.