Together AI
Together AI hosts 100+ open-source models with fast inference and competitive pricing. You get access to top models from Meta (Llama), DeepSeek, Qwen, and more through a single API key and OpenAI-compatible endpoint.
Getting an API Key
- Visit api.together.xyz/settings/api-keys
- Sign in or create a Together AI account
- Copy your API key
- Paste the key into AISCouncil under Settings > AI Model > Together AI
Together AI does not offer a free tier, but pricing is competitive. New accounts may receive starter credits. See together.ai/pricing for current rates.
API keys are stored locally in your browser (localStorage) and are never included in shared bot URLs.
Supported Models
| Model | Context | Max Output | Input Price | Output Price | Capabilities |
|---|---|---|---|---|---|
| Llama 3.3 70B Turbo | 128K | 4K | $0.88/MTok | $0.88/MTok | Tools, code, streaming |
| Llama 3.1 405B Turbo | 128K | 4K | $3.50/MTok | $3.50/MTok | Tools, code, streaming |
| DeepSeek R1 | 128K | 4K | $3.00/MTok | $7.00/MTok | Reasoning, code, streaming |
| Qwen 2.5 72B Turbo | 128K | 4K | $0.60/MTok | $0.60/MTok | Tools, streaming |
| Llama 4 Maverick | 1M | 4K | $0.24/MTok | $0.77/MTok | Vision, tools, code, streaming |
| Llama 4 Scout | 1M | 4K | $0.15/MTok | $0.40/MTok | Vision, tools, code, streaming |
Prices are per million tokens (MTok). See together.ai/pricing for the latest rates.
- Llama 4 Maverick offers 1M context with vision support at a low price -- great for multimodal tasks.
- Llama 3.3 70B Turbo is a solid all-rounder for general-purpose chat and coding at under $1/MTok.
- DeepSeek R1 provides strong reasoning capabilities for complex problem-solving.
100+ Open Models
Together AI hosts far more models than the curated list above. The models shown are the most popular ones pre-configured in AISCouncil. Together AI's catalog includes models from Mistral, Databricks, WizardLM, and many other providers.
Configuration
Select Together AI as the provider when creating a bot profile. The app connects to api.together.xyz/v1 using the OpenAI-compatible format with Bearer authentication.
Tips for Best Results
- Use Together AI to access open-source models at scale. If you want Llama, DeepSeek, or Qwen without running your own infrastructure, Together AI is a convenient option.
- Try Llama 4 Scout for budget-friendly 1M context. At $0.15/MTok input with a million-token window, it is one of the cheapest long-context options available.
- Compare models in a council. Use AISCouncil's compare mode to test Together AI models against each other or against closed models from other providers.