Use the right model for the task
Run suitable routine work on economical open models, and call on frontier models when deeper capability is worth the cost.
05 / Bring your own AI
Choose your AI.
Choose your cost.
Use supported models from OpenAI, Anthropic, Azure OpenAI, xAI, DeepSeek, Mistral and Groq, or open models such as Llama through Ollama, LM Studio or a compatible endpoint. Your memory and work stay in the same Cuttlefish workplace.
Included in the Cuttlefish workplace
What this gives you
No specialized GPU to buy
Hosted models and rented cloud computing can remove the need to purchase your own AI servers.
One workplace, not five subscriptions
Bring your model connections. Persistent memory, coordinated work, applications and audit history are part of the platform.
The value
Avoid paying premium model rates for every suitable task.
Move between supported providers as cost, performance and technology change, without rebuilding the work itself.
Supported Llama configurations can run on rented compute. A small-model setup at around $200 a month may be possible; capacity and suitability vary with the work.
Cuttlefish Founders
One workplace for the organization.
Cuttlefish is offered to organizations. Cuttlefish Founders is $499 USD a month including five members, plus $79 a month for each additional member. Larger managed Enterprise contracts are negotiated separately.
Consultants, MSPs and implementers: contact us about partner options.