Bring open models into your own network. Ogger combines LLMs with neuro-symbolic reasoning, retrieval and lasting memory to extend what your chosen model can do. Try it hosted, or run locally with control over your infrastructure and data.
Deploy Ogger with local inference in your own network. You choose the infrastructure, models and access controls. Using a cloud model sends inference requests to that provider; a fully local configuration keeps inference inside your environment.
Ogger combines neural language models with symbolic tools, evidence retrieval and memory. Our ambition is frontier-level results on selected tasks with open models. That is a goal to validate against your workloads, not a claim of general frontier-model parity.
from openai import OpenAI
client = OpenAI(
base_url="https://ogger.ai/api/v1",
api_key="og_live_...",
)
response = client.chat.completions.create(
model="ogger", # the engine: grounded, planning, remembers you
messages=[
{"role": "user", "content": "Refactor this codebase for production"}
],
stream=True,
)
for chunk in response:
print(chunk.choices[0].delta.content, end="")Works with your tools
Any tool that supports an OpenAI-compatible base URL works out of the box.
Enhance the model you choose with context, memory and tools. Results, latency and cost depend on the model and task.
Give your model retrieval, planning and symbolic calculations for tasks that benefit from evidence and exact computation.
Choose among supported open-source and open-weight models. See the selected model in chat and review its license; availability depends on your deployment.
Keep project requirements and preferences across conversations so your model can use the context you have already shared.
Use supported local models or cloud-hosted open models. Start with the hosted experience and configure inference for your own deployment.
Inspect reported token usage and estimated cost per query. Compare the added reasoning work with the result it produces.
Connect through OpenAI-compatible and Anthropic-compatible APIs, with API keys managed from your dashboard.