# qwen/qwen2.5-vl-72b-instruct

Calls to the model qwen/qwen2.5-vl-72b-instruct on Nebius stopped working on August 31, 2026.

Nebius retired model qwen/qwen2.5-vl-72b-instruct on August 31, 2026. Calls specifying this model fail since that date. The retirement affects requests sent to the host api.studio.nebius.ai.

## What to use instead

Move to MiniMaxAI/MiniMax-M3, which is the replacement recommended by Nebius.

- API: Nebius (api.studio.nebius.ai)
- Model: qwen/qwen2.5-vl-72b-instruct
- Released: January 26, 2025 (models.dev)
- Kind: Model retirement
- Takes effect: August 31, 2026
- Calls after that day: fail
- Calls it hits: `model qwen/qwen2.5-vl-72b-instruct`
- Source: [Model notices](https://docs.tokenfactory.nebius.com/llms.txt#deprecation)
- Verified: November 6, 2026, against Nebius's own pages
- Page: https://www.okapi.bot/models/nebius/qwen/qwen2.5-vl-72b-instruct/

## Nebius's words

> Deprecating Model: Qwen/Qwen2.5-VL-72B-Instruct; Recommended Replacement: MiniMaxAI/MiniMax-M3

Okapi finds the calls this change hits in your repo and opens the pull request with the fix: https://github.com/apps/okapi-agent