LLM Reference
Cohere API

North Mini Code 1.0 on Cohere API

North · Cohere

ServerlessOpen Source

Last refreshed 2026-06-09. Next refresh: weekly.

Why use North Mini Code 1.0 on Cohere API?

Cohere API offers North Mini Code 1.0 with competitive pricing. Cohere is a leading enterprise AI company that specializes in developing large language models (LLMs) and Retrieval-Augmented Generation (RAG) capabilities.

Input / 1M
-
Output / 1M
-
Cache
Not sourced
Batch
Not sourced

Setup recipe

Docs fallback
Install
Use the provider REST API or SDK
Auth
Create a provider API key
Call
model: north-mini-code-1-0
Model ID
north-mini-code-1-0

Request example

Curated snippets for this provider are not sourced yet. Use Cohere API documentation with model ID north-mini-code-1-0.

Gotchas

  • Use provider model ID "north-mini-code-1-0", not the LLMReference slug "north-mini-code-1.0".

Capabilities

ReasoningFunction CallingTool UseStructured Outputs

About North Mini Code 1.0

A 30B-parameter sparse mixture-of-experts (3B active parameters) code generation model from Cohere, released as open weights under Apache 2.0. Designed for agentic software engineering, code generation, and terminal-based tasks. Supports a 256K token context window with 64K maximum output length, interleaved thinking, and tool use via JSON schema chat templates. The inaugural model in Cohere's North family.

FAQ

What is the context window for North Mini Code 1.0 on Cohere API?

North Mini Code 1.0 supports a 256k token context window on Cohere API.

What API model ID do I use for North Mini Code 1.0 on Cohere API?

Use the model ID north-mini-code-1-0 when calling Cohere API's API.

Who created North Mini Code 1.0?

North Mini Code 1.0 was created by Cohere as part of the North model family.

Is North Mini Code 1.0 open source?

North Mini Code 1.0 is open source under Apache 2.0 according to the seed data.

Get Started

Model Specs

Released2026-06-09
Parameters30B (MoE, 3B active)
Context256k
ArchitectureDecoder-only Transformer, sparse mixture-of-experts (128 experts, 8 active per token), interleaved sliding-window attention with RoPE and global attention in 3:1 ratio, SwiGLU FFN with sigmoid-gated routing