AI & Machine Learning
Connect Together AI to Your AI Agents
Let your Wkil agents run prompts on hosted open models, fine-tune custom models, and execute code in a sandbox through Together AI.
Authentication: API key
What you can do
- Run prompts against hosted open models
- Fine-tune and deploy custom models
- Execute generated code in a sandbox
Available capabilities may vary by authentication and workspace configuration.
Actions supported by the connector
Run an AI model
Send a prompt to a hosted open model and get a completion.
Fine-tune a model
Train a custom fine-tuned version of an open model.
Launch a GPU cluster
Provision GPU compute for training or inference workloads.
Use the evals API
Run automated evaluations against a model's outputs.
Run code in a sandbox
Execute generated code safely in an isolated environment.
Deploy dedicated containers
Deploy a model to dedicated, always-on infrastructure.
Bring your own model
Upload and serve a custom model on the platform.
Practical agent use cases
- An agent runs a prompt against an open model to draft a response.
- An agent executes generated code in a sandbox before returning the result.
How it works with Wkil
- 1Choose the integrationAdd Together AI to your agent from inside the Wkil platform.
- 2Grant the permissions it needsDefine precisely what the agent can access and perform.
- 3Test, then run itTest actions before they go live, with human approval where needed.
Permissions always stay in your control — you can scope access and require human approval before any sensitive action.
Related integrations
FAQ
Can an agent use a custom fine-tuned model?
Yes — Together AI supports fine-tuning and deploying your own custom models.