# TuningEngines

> TuningEngines is the Universal Intelligence Runtime — one drop-in OpenAI-compatible API for open, commercial, and fine-tuned models, with built-in tuning, evaluations, and deployment. Built so AI compounds with your business.

TuningEngines (by Ockham Labs Inc., platform: CerebrixOS) lets teams call any model — open-source, commercial, or privately fine-tuned — through a single endpoint that is fully OpenAI SDK compatible. We host open models, host your fine-tuned models, run evaluations, and support the full LiteLLM feature surface (routing, fallbacks, budgets, observability).

## Quick facts

- **Endpoint**: `https://api.tuningengines.com/v1`
- **Compatibility**: OpenAI Python/JS SDKs, curl, LangChain, LlamaIndex, any OpenAI-compatible client
- **Auth**: Bearer token, keys formatted as `sk-te-…`
- **Capabilities**: Chat completions, embeddings, vision, function/tool calling, streaming, fine-tuning, hosting of custom adapters
- **Deployment tiers**: Managed, Hybrid, Sovereign (on-prem / private cloud)
- **Positioning**: "AI that compounds with your business"

## Example usage

```bash
curl https://api.tuningengines.com/v1/chat/completions \
  -H "Authorization: Bearer sk-te-..." \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Hello"}]}'
```

```python
from openai import OpenAI
client = OpenAI(base_url="https://api.tuningengines.com/v1", api_key="sk-te-...")
client.chat.completions.create(model="gpt-4o-mini", messages=[{"role":"user","content":"Hello"}])
```

## Pages

- [Home](https://tuningengines.com/): Universal Intelligence Runtime, unified API, model library, lifecycle, use cases
- [About](https://tuningengines.com/about): Company, team (CEO, CTO), mission
- [Contact](https://tuningengines.com/contact): Sales, support, partnerships
- [Open Source](https://tuningengines.com/open-source): Shared Source Initiative — open domain-specific models and datasets
- [Privacy Policy](https://tuningengines.com/privacy-policy): Data handling and sovereignty commitments

## Product surface

- **Unified API**: One OpenAI-compatible API for any model. Drop-in replacement for `openai.base_url`.
- **Model Library**: Hosted open models (Llama, Mistral, Qwen, DeepSeek families and more) plus commercial frontier models, all addressable by name.
- **Fine-tuning & Hosting**: Train adapters on your data; we host them behind the same API with per-token billing.
- **Lifecycle**: Tune → Evaluate → Deploy → Observe → Iterate.
- **LiteLLM-grade routing**: Fallbacks, retries, budgets, rate limits, virtual keys, cost tracking.
- **Sovereign deployments**: Run the full stack inside your VPC or on-prem for regulated workloads.

## Who it's for

Engineering teams shipping copilots, agents, RAG systems, structured-extraction pipelines, and regulated decision-support — who need model choice, cost control, evaluations, and the option to bring intelligence in-house.

## Contact

- Website: https://tuningengines.com
- App / Console: https://app.tuningengines.com
- Legal entity: Ockham Labs Inc.
- Platform: CerebrixOS