Together AI
meta-llama/llama-guard-4-12b model retirement
Stopped workingSubscribe to Together AIShare
Calls to the model meta-llama/llama-guard-4-12b on Together AI stopped working on August 25, 2026.
In Together AI's words
Removal date: 2026-08-25; Model: meta-llama/Llama-Guard-4-12B; Supported by on-demand dedicated endpoints: No
Facts
- API
- Together AI
api.together.xyz - Model
meta-llama/llama-guard-4-12b- Released
- April 29, 2025 (models.dev)
- Kind
- Model retirement
- Takes effect
- August 25, 2026
- Calls after that day
- Fail
- Source
- Deprecations page
- Last updated
- Verified
- , against Together AI's own pages
Questions
When does meta-llama/llama-guard-4-12b shut down?
Calls to the model meta-llama/llama-guard-4-12b on Together AI stopped working on August 25, 2026.
What happens to calls after August 25, 2026?
They have failed since August 25, 2026.
When was meta-llama/llama-guard-4-12b released?
meta-llama/llama-guard-4-12b was released on April 29, 2025, 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-guard-4-12b on other APIs 1
-
Groq
meta-llama/llama-guard-4-12bModel retirement Stopped working
More from Together AI 10
-
black-forest-labs/flux.1-schnellModel retirement Stopped working -
moonshotai/kimi-k2.6Model retirement Stopped working -
deepcogito/cogito-v2-1-671bModel retirement Stopped working -
google/gemma-3n-e4b-itModel retirement Stopped working -
deepseek-ai/deepseek-v4-proModel retirement Stopped working -
moonshotai/kimi-k2.7-codeModel retirement Stopped working -
nvidia/nemotron-3-ultra-550b-a55bModel retirement Stopped working -
pearl-ai/gemma-4-31b-itModel retirement Stopped working -
nvidia/nemotron-3-nano-omni-30b-a3b-reasoning-bf16Model retirement Stopped working -
nvidia/nvidia-nemotron-3-super-120b-a12b-bf16Model retirement Stopped working
You fixed this one by hand.
Okapi opens the pull request with the fix before the next breaking change reaches your code.