Models
Relai provides access to frontier models from leading providers through private cloud deployments on Amazon Bedrock and Azure OpenAI. All models are available in both EU and US regions.
Private Cloud Inference
Unlike public API endpoints, Relai routes your requests through private cloud deployments on Amazon Bedrock and Azure:
- Zero training on your data — Frontier labs (Anthropic, OpenAI, Google) never see or train on your prompts
- No logging by model providers — Your conversations are not stored or analyzed by the AI companies
- Enterprise-grade isolation — Dedicated compute with SOC2 and HIPAA compliance
- Contractual guarantees — Data processing agreements with AWS and Microsoft, not frontier labs
Pricing
Transparent pricing with no markup. You pay the same rates as direct Bedrock/Azure access.
| Model | Provider | Input (1M tokens) | Output (1M tokens) | Context |
|---|---|---|---|---|
| claude-opus-4.7 | Anthropic | $15.00 | $75.00 | 200k |
| claude-sonnet-4.6 | Anthropic | $3.00 | $15.00 | 200k |
| claude-haiku-4 | Anthropic | $0.80 | $4.00 | 200k |
| gpt-5.5 | OpenAI | $5.00 | $15.00 | 256k |
| gpt-5.5-mini | OpenAI | $0.60 | $2.40 | 256k |
| o4-mini | OpenAI | $1.10 | $4.40 | 200k |
| gemini-2.5-pro | $2.50 | $10.00 | 2M | |
| gemini-2.5-flash | $0.15 | $0.60 | 1M | |
| llama-4-maverick | Meta | $0.20 | $0.60 | 1M |
| llama-4-scout | Meta | $0.17 | $0.85 | 512k |
| mistral-large-3 | Mistral | $2.00 | $6.00 | 128k |
Model IDs
Use the full model ID in your API requests:
{
"model": "claude-sonnet-4.6",
"messages": [...]
}Provider-Prefixed Models
You can also use provider-prefixed model IDs:
{
"model": "anthropic/claude-sonnet-4.6",
"messages": [...]
}This is useful when a model name exists across multiple providers.
Model Capabilities
| Model | Vision | Tools | JSON Mode | Streaming |
|---|---|---|---|---|
| claude-opus-4.7 | ✓ | ✓ | ✓ | ✓ |
| claude-sonnet-4.6 | ✓ | ✓ | ✓ | ✓ |
| claude-haiku-4 | ✓ | ✓ | ✓ | ✓ |
| gpt-5.5 | ✓ | ✓ | ✓ | ✓ |
| gpt-5.5-mini | ✓ | ✓ | ✓ | ✓ |
| o4-mini | ✓ | ✓ | ✓ | ✓ |
| gemini-2.5-pro | ✓ | ✓ | ✓ | ✓ |
| gemini-2.5-flash | ✓ | ✓ | ✓ | ✓ |
| llama-4-maverick | ✓ | ✓ | ✓ | ✓ |
| llama-4-scout | ✗ | ✓ | ✓ | ✓ |
| mistral-large-3 | ✓ | ✓ | ✓ | ✓ |
Infrastructure
All inference runs on:
- Amazon Bedrock — Claude models (Anthropic), Llama models (Meta), Mistral models
- Azure OpenAI — GPT and O-series models (OpenAI)
This ensures your data stays within enterprise cloud boundaries and is never exposed to frontier lab training pipelines.
EU-Safe Models
All models accessible through the EU region (eu.api.llmrelai.com) route through EU-based Bedrock and Azure deployments. Your prompts and completions never leave the EU.
For a complete list of models, see the pricing page.