# Codestral Mamba

DevFeed: [Codestral Mamba](<https://devfeed.tech/articles/codestral-mamba-6988.md>)

Original publisher: [Read original article](<https://mistral.ai/news/codestral-mamba/>)

Published: 2024-07-16T08:00:00Z

Content type: news

Language: en

Sources: [Mistral AI Blog](<https://devfeed.tech/sources/mistral-ai-blog.md>)

Topics: [Mamba](<https://devfeed.tech/topics/mamba.md>), [code productivity](<https://devfeed.tech/topics/code-productivity.md>), [Inference](<https://devfeed.tech/topics/inference.md>), [Local AI](<https://devfeed.tech/topics/local-ai.md>), [SDK](<https://devfeed.tech/topics/sdk.md>), [TensorRT-LLM](<https://devfeed.tech/topics/tensorrt-llm.md>), [llama.cpp](<https://devfeed.tech/topics/llama-cpp.md>)

Tags: [efficiency](<https://devfeed.tech/tags/efficiency.md>), [inference](<https://devfeed.tech/tags/inference.md>), [llama-cpp](<https://devfeed.tech/tags/llama-cpp.md>), [mamba](<https://devfeed.tech/tags/mamba.md>), [mixtral](<https://devfeed.tech/tags/mixtral.md>), [open](<https://devfeed.tech/tags/open.md>), [sdk](<https://devfeed.tech/tags/sdk.md>), [tensorrt-llm](<https://devfeed.tech/tags/tensorrt-llm.md>)

## AI overview

Mistral introduces Codestral Mamba, a freely usable and distributable code-focused model based on the Mamba architecture. It offers linear-time inference, supports very long sequences, and is intended for efficient local code-assistant use. The article describes deployment through the mistral-inference SDK and TensorRT-LLM, with anticipated llama.cpp support, and notes its Apache 2.0 license.

## Source excerpt

The most powerful AI platform for enterprises. Customize, fine-tune, and deploy AI assistants, autonomous agents, and multimodal AI with open models.