Nebius
meta-llama/llama-3.3-70b-instruct model retirement
Stopped workingSubscribe to NebiusShare
On August 31, 2026, Nebius retired model meta-llama/llama-3.3-70b-instruct. Requests using this model identifier fail since that date.
What to use instead of meta-llama/llama-3.3-70b-instruct
Move to nvidia/Nemotron-3_5-Lightning, which is recommended by Nebius.
In Nebius's words
Deprecating Model: meta-llama/Llama-3.3-70B-Instruct; Recommended Replacement: nvidia/Nemotron-3_5-Lightning
Facts
- API
- Nebius
api.studio.nebius.ai - Model
meta-llama/llama-3.3-70b-instruct- Released
- December 6, 2024 (models.dev)
- Kind
- Model retirement
- Takes effect
- August 31, 2026
- Calls after that day
- Fail
- Source
- Model notices
- Last updated
- Verified
- , against Nebius's own pages
Questions
When does meta-llama/llama-3.3-70b-instruct shut down?
Calls to the model meta-llama/llama-3.3-70b-instruct on Nebius stopped working on August 31, 2026.
What happens to calls after August 31, 2026?
They have failed since August 31, 2026.
When was meta-llama/llama-3.3-70b-instruct released?
meta-llama/llama-3.3-70b-instruct was released on December 6, 2024, according to models.dev.
How do I find the calls this hits in my code?
Okapi reads your repository, finds the calls this change hits and opens a pull request with the fix. It never commits to your branches: every change comes as a pull request you review.
meta-llama/llama-3.3-70b-instruct on other APIs 3
-
DeepInfra
meta-llama/llama-3.3-70b-instructModel retirement Stopped working -
Novita AI
meta-llama/llama-3.3-70b-instructModel retirement Stops workingin 20 days -
Azure AI Foundry
llama-3.3-70b-instructModel retirement Stops workingin 86 days
More from Nebius 10
-
deepseek-ai/deepseek-v4-flashModel retirement Stopped working -
minimaxai/minimax-m2.5Model retirement Stopped working -
nousresearch/hermes-4-70bModel retirement Stopped working -
nvidia/cosmos3-super-reasonerModel retirement Stopped working -
nvidia/llama-3_1-nemotron-ultra-253b-v1Model retirement Stopped working -
nvidia/nemotron-3-nano-omniModel retirement Stopped working -
qwen/qwen2.5-vl-72b-instructModel retirement Stopped working -
qwen/qwen3-32bModel retirement Stopped working -
qwen/qwen3-next-80b-a3b-thinkingModel retirement Stopped working -
deepseek-ai/deepseek-v4-flash-0731Model retirement Stops workingin 3 days
You fixed this one by hand.
Okapi opens the pull request with the fix before the next breaking change reaches your code.