# Modal + Mistral 3: 10x faster cold starts with GPU snapshotting

DevFeed: [Modal + Mistral 3: 10x faster cold starts with GPU snapshotting](<https://devfeed.tech/articles/modal-mistral-3-10x-faster-cold-starts-with-gpu-snapshotting-79884.md>)

Original publisher: [Read original article](<https://modal.com/blog/mistral-3>)

Author: Modal

Published: 2025-12-02T12:00:00Z

Content type: article

Language: en

Sources: [Modal](<https://devfeed.tech/sources/modal.md>)

Topics: [tgi](<https://devfeed.tech/topics/tgi.md>), [LLM optimization](<https://devfeed.tech/topics/llm-optimization.md>), [mixtral](<https://devfeed.tech/topics/mixtral.md>)

Tags: [day-0-support](<https://devfeed.tech/tags/day-0-support.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [gpu](<https://devfeed.tech/tags/gpu.md>), [llm-serving](<https://devfeed.tech/tags/llm-serving.md>), [mistral](<https://devfeed.tech/tags/mistral.md>), [modal](<https://devfeed.tech/tags/modal.md>), [performance](<https://devfeed.tech/tags/performance.md>), [snapshot](<https://devfeed.tech/tags/snapshot.md>)

## AI overview

Modal announces day-zero support for Mistral 3 and provides a vLLM deployment example for Ministral 3. Its serverless GPU infrastructure scales with demand, and GPU snapshots reduced median cold-start time for the tested 3B model from about 118 seconds to about 12 seconds.

## Source excerpt

We've partnered with Mistral to bring you Day 0 support for Mistral 3, with GPU-snapshot-optimized performance.