Together AI Provider

Together AI hosts open-source models with competitive pricing and a free tier. They offer a wide catalog of Llama, Mixtral, Qwen, and other open models.

Get an API Key

Go to api.together.ai
Create an account or sign in
Navigate to API Keys
Create a new key and copy it

Together AI offers a free tier with limited usage.

Configure VibeCody

Option 1: Environment variable (recommended)

export TOGETHER_API_KEY="..."
vibecli --provider together

Option 2: Config file (~/.vibecli/config.toml)

[together]
enabled = true
api_key = "..."
model = "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo"

Model Selection

Model	Strengths	Best for
`meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo`	Strong general coding	Daily tasks (default)
`meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo`	Ultra-fast	Quick completions
`Qwen/Qwen2.5-Coder-32B-Instruct`	Code-specialized	Code generation
`mistralai/Mixtral-8x22B-Instruct-v0.1`	Large context	Longer analysis

Default: meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo

Override from the CLI:

vibecli --provider together --model "Qwen/Qwen2.5-Coder-32B-Instruct"

Browse all models at api.together.ai/models.

Pricing

Together AI offers competitive pricing for open models. Free tier includes limited tokens per day. Check together.ai/pricing for current rates.

Best For

Open-source models – access the latest Llama, Qwen, Mixtral releases
Cost-effective – lower prices than running your own GPU infrastructure
Free tier – experiment with models before committing
Model variety – large catalog of open models

Verify Connection

vibecli --provider together -c "Write a Rust function to parse command-line arguments"

Provider: Together AI