Learn about the models available on June, how they're served, and what this means for your privacy.
Available Models
Retrieve the full list of models available to you via the /v1/models endpoint.
curl https://api.blockchain.info/ai/api/v1/models \
-H "Authorization: Bearer $JUNE_API_KEY"
Open Weights vs Proprietary Models
June offers two categories of model: open weights and proprietary.
Open Weights Models
Open weights models (such as Llama, Mistral, DeepSeek, and Qwen) have publicly available model weights. June serves these through dedicated inference providers that guarantee no prompts are stored or used for training. Because the weights are open, we can select providers that meet our strict privacy requirements while still offering high-quality inference.
Proprietary Models
Proprietary models (such as those from OpenAI and Anthropic) are served by their respective providers. Your prompts are sent to the provider for processing. These providers have their own privacy policies, but June's API terms ensure that prompts sent through our platform are not used for model training.
Privacy & Anonymity
Regardless of which model you use, June does not send any user-identifying data with your prompts. June acts as an intermediary that strips identifying information before your prompt reaches any provider, so your requests are indistinguishable from those of any other June user.
Model Aliases
To save you from updating your code every time we release a new model version or retire an older one, June provides stable aliases that always route to the latest version of each model family that we support. Use these in place of version-specific slugs to keep your integration working without changes as the model landscape evolves.
anthropic/claude-opus-latestanthropic/claude-sonnet-latestanthropic/claude-haiku-latestgoogle/gemini-pro-latestgoogle/gemini-flash-latestopenai/gpt-latestopenai/gpt-mini-latestx-ai/grok-latest
If you need to pin to a specific model version, use the versioned slug directly instead.