What's the price range for Cohere models on OpenKey?

From free (North Mini Code) up to $2.50 input / $10 output per 1M tokens for Command A and Command R+ (08-2024). On OpenKey, provider list price gets a flat 3% fee, so Command A comes to $2.50 x 1.03 = $2.575 input and $10 x 1.03 = $10.30 output per 1M tokens.

Which Cohere model should I use if I want the cheapest paid option?

Command R7B (12-2024) at $0.0375 input / $0.15 output per 1M tokens is the cheapest non-free Cohere model, with a 128K context window. It's a reasonable choice for high-volume tasks where Command A's cost isn't justified.

Is there a free Cohere model on OpenKey?

Yes — North Mini Code (free), released 2026-06-17, with a 256K context window and zero cost on input and output. It's one of 25 free models across OpenKey's full 329-model roster.

Cohere

5 models built · 5 served

Cohere is a Canadian AI lab founded in 2019, built around the Command model family and positioned toward enterprise use, particularly retrieval-augmented generation. On OpenKey you get 5 Cohere models: the flagship Command A, the previous-gen Command R and Command R+ (both dated 08-2024), the smaller Command R7B, and a free coding model, North Mini Code. Cohere's lineup strategy is narrow and deliberate rather than sprawling — a handful of well-defined tiers instead of dozens of variants, each targeting a different cost/capability point for production RAG and enterprise text workloads. All five models are text-to-text, and most support large 128K-256K context windows.

Lineup

Command A ($2.50/$10 per 1M tokens) is the flagship, sharing Command R+'s pricing but with a newer March 2025 build and 256K context — use it for the hardest tasks. Command R+ (08-2024) sits at the same price with 128K context, effectively superseded by Command A. For cheaper production work, Command R (08-2024) runs $0.15/$0.60 per 1M with 128K context — a solid middle tier. Command R7B is the budget pick at $0.0375/$0.15 per 1M. If you just need to prototype for free, North Mini Code (free) has zero cost and a 256K context window.

Model Context Input /1M Output /1M

North Mini Code (free)FreeJun 17, 2026 256K Free Free ctx 256K · in Free · out Free
Command AMar 13, 2025 256K $2.58 $10.30 ctx 256K · in $2.58 · out $10.30
Command R7B (12-2024)Dec 14, 2024 128K $0.039 $0.154 ctx 128K · in $0.039 · out $0.154
Command R (08-2024)Aug 30, 2024 128K $0.154 $0.618 ctx 128K · in $0.154 · out $0.618
Command R+ (08-2024)Aug 30, 2024 128K $2.58 $10.30 ctx 128K · in $2.58 · out $10.30

Prices per 1M tokens, flat 3% fee included.

Models Cohere serves

Beyond its own models, Cohere hosts 5 models in the OpenKey catalog. Prices are Cohere’s own list price per 1M input tokens, before our fee.

Model	Built by	Context	Input /M
Command A	Cohere	256K	$2.50
Command R (08-2024)	Cohere	128K	$0.150
Command R+ (08-2024)	Cohere	128K	$2.50
Command R7B (12-2024)	Cohere	128K	$0.037
North Mini Code (free)	Cohere	256K	Free

From the OpenRouter catalog, Jul 28, 2026. All providers

Status page ↗Privacy policy ↗Terms of service ↗

Questions

What's the price range for Cohere models on OpenKey?: From free (North Mini Code) up to $2.50 input / $10 output per 1M tokens for Command A and Command R+ (08-2024). On OpenKey, provider list price gets a flat 3% fee, so Command A comes to $2.50 x 1.03 = $2.575 input and $10 x 1.03 = $10.30 output per 1M tokens.
Which Cohere model should I use if I want the cheapest paid option?: Command R7B (12-2024) at $0.0375 input / $0.15 output per 1M tokens is the cheapest non-free Cohere model, with a 128K context window. It's a reasonable choice for high-volume tasks where Command A's cost isn't justified.
Is there a free Cohere model on OpenKey?: Yes — North Mini Code (free), released 2026-06-17, with a 256K context window and zero cost on input and output. It's one of 25 free models across OpenKey's full 329-model roster.