Ottili AI PlatformOttili AI

Sign in Start building

OpenAI-compatible · usage-based

Every AI workload.
One intelligent platform.

One OpenAI-compatible API for Ottili's own models. Build with Vale, go deeper with Cairn, or let Auto route between them — with usage controls, prepaid billing and API keys built in.

Start building Explore models

Built for the Ottili ecosystem

Ottili ONE Ottili Coder Ottili LD3 Ottili Cloud

OpenAI-compatible APILive model routingPrepaid usage controlsScoped API keys

API status: Operational

Unified AI Platform

Stay focused.
Stay connected.

Ottili Vale and Ottili Cairn, an adaptive Auto router, and every Ottili product surface that needs inference — all reachable through one OpenAI-compatible endpoint, one key system, one usage ledger.

3: Production models live today
1M: Token context window per model
100%: OpenAI-compatible request format
1: Unified API for every model

One unified AI platform

One API for every Ottili model

OpenAI-compatible /v1/chat/completions and /v1/responses
One base URL and one key format across Vale, Cairn and Auto
Centralized usage, billing and key management

Read the API docs

Execution that adapts

Routing built to fail closed, not silently

Workload-aware selection between Vale and Cairn via Auto
Provider health checks before every dispatch
Fallback only when explicitly permitted per model and company

View live status

Better intelligence per credit

Transparent, usage-based cost control

Prepaid credits — a request is rejected up front if balance can't cover it
Per-model, per-token pricing shown before you spend
Vale for high-volume work, Cairn for depth, Auto to blend both

See pricing & credit packs

Control at every layer

Scoped keys, isolated by company

otk_live_ / otk_test_ API keys with per-key model and budget limits
Every request resolved to a single company — no cross-tenant reads
Usage limits and rate limiting enforced per key and per company

Manage API keys

Built for real execution

Ottili AI Platform isn't a side project bolted onto the ecosystem — it's the model execution layer other Ottili surfaces already build on.

Ottili Coder Cloud runs on Ottili AI Platform
Ottili Coder's cloud coding surface uses Vale, Cairn and Auto as its native model provider — not a third-party reseller layer.
Shared credits and API keys
One prepaid balance and one key system across every surface that calls Ottili's models.
One platform, not a bolt-on API
Built inside Ottili ONE from the start — company-scoped, auditable, and versioned alongside the rest of the platform.

Ottili AI integrations

One AI surface.
Every integration on display.

Ottili AI is the AI surface inside Ottili ONE. Every capability below — Vale, Cairn, adaptive routing, missions, DWE specialists, tools, approvals, memory, credits, cost control and audit — is reachable through one OpenAI-compatible API and shares one company context, one credit ledger and one audit trail with Ottili HQ, Ottili Coder, Ottili LD3 and Ottili Cloud.

Available

Ottili Vale

Integration with Ottili Ai — wired through the Ottili AI surface.

Available

Ottili Cairn

Integration with Ottili Ai — wired through the Ottili AI surface.

Available

Adaptive routing (Ottili Auto)

Integration with Ottili Ai — wired through the Ottili AI surface.

Available

Missions

Integration with Ottili Ai — wired through the Ottili AI surface.

Available

DWE specialists

Integration with Otili-Ai — wired through the Ottili AI surface.

Available

Tool use

Integration with Ottili Ai — wired through the Ottili AI surface.

Available

Approvals

Integration with Ottili Console — wired through the Ottili AI surface.

Planned

Memory

Integration with Ottili Ai — wired through the Ottili AI surface.

Available

Credits

Integration with Ottili Core — wired through the Ottili AI surface.

Available

Cost control

Integration with Ottili Core — wired through the Ottili AI surface.

Available

Audit

Integration with Ottili Core — wired through the Ottili AI surface.

Available

Native to Ottili ONE

Integration with Ottili One — wired through the Ottili AI surface.

Explore the Ottili AI docs

Ottili AI architecture

One AI surface.
Every capability on display.

From one request to a logged, cost-controlled result — the AI surface of Ottili ONE.

Available

Ottili Vale

Part of Ottili ai.

Available

Ottili Cairn

Part of Ottili ai.

Available

Adaptive routing (Ottili Auto)

Part of Ottili ai.

Available

Missions

Part of Ottili ai.

Available

DWE specialists

Part of otili-ai.

Available

Tool use

Part of Ottili ai.

Available

Approvals

Part of Ottili console.

Planned

Memory

Part of Ottili ai.

Available

Credits

Part of Ottili core.

Available

Cost control

Part of Ottili core.

Available

Audit

Part of Ottili core.

Available

Native to Ottili ONE

Part of Ottili one.

FAQ

Questions about Ottili AI

Common questions about how the Ottili AI public surface works — models, routing, missions, specialists, tools, approvals, memory, credits, audit and cost control — and how it plugs into the rest of Ottili ONE.

What is Ottili AI?: Ottili AI is the AI control surface and assistant layer of Ottili ONE. It runs Ottili's own models — Vale and Cairn — behind one OpenAI-compatible API, with adaptive routing, missions, the Dynamic Workforce Engine (DWE), tool use and usage controls. It powers Ottili Coder, Ottili LD3 and other Ottili surfaces from inside your company context.
Which models are available?: Ottili Vale (efficient, everyday work) and Ottili Cairn (premium, deeper reasoning) are live today, plus Ottili Auto, which adaptively routes each request to the right model. The per-model status on this page reflects the canonical product-status registry — nothing here claims an unverified model.
How does routing work?: Ottili Auto adaptively routes each request: everyday fixes go to Vale, heavier engineering and reasoning go to Cairn. Routing is part of the Ottili AI Platform and is callable through the same OpenAI-compatible endpoint — you change only the model id, not your code.
What are missions and DWE specialists?: A mission is a goal you hand to Ottili AI; the Dynamic Workforce Engine (DWE) breaks it into work and assigns DWE specialists — role-based agents that carry the work forward. Missions and DWE are part of the Ottili AI capability set shown on this page; their readiness is taken from the canonical product-status registry.
How are tools and approvals handled?: DWE specialists can call tools (the Unified API business.* and ld3.* namespaces) to act on your data. Critical actions are approval-first: nothing leaves your company context for an external effect until a required approval is granted, enforced through Ottili Console and Ottili Core.
Does Ottili AI remember context?: Memory is part of the Ottili AI capability set. The public readiness shown on this page is the canonical product-status value (currently planned until enforced) — we do not claim a memory capability that is not yet verified as live.
How do credits and cost control work?: Ottili AI is prepaid: you add credits, and each request's cost is reserved and deducted from your balance on real per-model token pricing. A request is rejected up front if the balance cannot cover it. Keys support per-key model and budget limits, and rate limiting is enforced per key and per company — that is the cost-control surface demonstrated here.
How is usage audited?: Every AI action is resolved to a single company and recorded through Ottili Core audit, so usage, approvals and tool calls are traceable inside your company boundary. Audit is a Core-backed capability shown on this page.
How does Ottili AI connect to the rest of Ottili ONE?: Ottili AI is native to Ottili ONE: it shares one login, one company context, one credit ledger and one audit trail with Ottili HQ, Ottili Coder, Ottili LD3 and Ottili Cloud. The same models and routing power the other surfaces from inside your company.

Documentation

Read the Ottili AI docs

Go deeper on each part of the Ottili AI surface. These guides explain how models, routing, missions, DWE specialists, tools, approvals, memory, credits, audit and cost control work in Ottili AI. Every link points to the public Ottili documentation; nothing here claims an availability the docs do not state.

ModelsOttili Vale, Ottili Cairn and the model catalog.Read docs RoutingHow Ottili Auto routes requests between models.Read docs MissionsHand goals to Ottili AI and track them as missions.Read docs DWE specialistsThe Dynamic Workforce Engine and its specialists.Read docs ToolsTool use across the Unified API business and ld3 namespaces.Read docs ApprovalsApproval-first AI actions through Ottili Console.Read docs CreditsPrepaid credits and per-request cost reservation.Read docs Cost controlPer-key model and budget limits and rate limiting.Read docs AuditCompany-scoped audit of AI usage and approvals.Read docs

See it run

One request, end to end.
The Ottili AI surface.

Step through how a single request moves through Ottili AI — Vale and Cairn, adaptive routing, missions, DWE specialists, tool use, approvals, memory, credits, cost control and audit — native to Ottili ONE. Every step shows its honest registry readiness.

Ottili AIExample

Conceptual preview · example data

How Ottili AI runs a workload

From one request to a logged, cost-controlled result — the AI surface of Ottili ONE.

Ottili Vale is a real Ottili AI capability in the request flow shown here, with its honest registry status.
Capability type
model
Readiness
Available
via
Ottili Ai

Conceptual preview · example data

Pricing & availability

What Ottili AI offers today.

Every capability below is part of the Ottili AI surface — Vale, Cairn, adaptive routing, missions, DWE specialists, tools, approvals, memory, credits, cost control and audit. Availability comes from the Ottili product registry — nothing is shown as available that is not. Pricing is sourced from the central Ottili ONE pricing catalog; no price is shown that the catalog does not attest.

Availability

Available
Ottili Valevia Ottili Ai
Available
Ottili Cairnvia Ottili Ai
Available
Adaptive routing (Ottili Auto)via Ottili Ai
Available
Missionsvia Ottili Ai
Available
DWE specialistsvia Otili-Ai
Available
Tool usevia Ottili Ai
Available
Approvalsvia Ottili Console
Planned
Memoryvia Ottili Ai
Available
Creditsvia Ottili Core
Available
Cost controlvia Ottili Core
Available
Auditvia Ottili Core
Available
Native to Ottili ONEvia Ottili One

Pricing

Ottili AI pricing is not published as a standalone plan yet. Ottili AI runs on prepaid credits billed per request on real per-model token pricing (Vale, Cairn, Auto), with cost control enforced per key and per company. Plan prices come from the central Ottili ONE pricing catalog; AI plan packaging is added here only when grounded in that catalog.

Security & data handling

How Ottili AI handles your data.

Every Ottili AI workload runs inside a single company boundary: requests are routed and resolved to one company, run by Vale and Cairn with DWE specialists that call tools only with approvals, paid for by prepaid credits with cost control, and recorded by audit. Availability comes from the Ottili product registry — nothing is shown as available that is not.

Available
Company-scoped isolationvia Ottili Core
Available
Routing inside your contextvia Ottili Ai
Available
Approval-first toolsvia Ottili Console
Planned
Memory boundaryvia Ottili Ai
Available
Prepaid creditsvia Ottili Core
Available
Cost control & rate limitsvia Ottili Core
Available
Audit trailvia Ottili Core
Available
Ottili Valevia Ottili Ai
Available
Ottili Cairnvia Ottili Ai
Available
Missionsvia Ottili Ai
Available
DWE specialistsvia Ottili Ai
Available
Tool usevia Ottili Ai
Available
Native to Ottili ONEvia Ottili One

“Ottili AI Platform powers the model workloads behind Ottili Coder's cloud coding surface — Vale and Cairn run real, production coding requests today, not a demo.”

Ottili Coder — Ottili ecosystem integration, not an external customer

1M

Token context window

100%

OpenAI-compatible

Ottili models

Every model.
One catalog.

Full model catalog & specs

Live

Ottili Vale 1.2

Efficient tier for everyday developer and business work — fast and cost-effective.

$0.20/$0.40 per 1M1000K ctx

Live

Ottili Cairn 1.2

Premium price-performance tier for serious software and analysis work.

$0.65/$1.30 per 1M1000K ctx

Live

Ottili Auto

Adaptive router that picks Vale or Cairn per request — Vale for everyday work, Cairn for serious engineering.

$0.43/$0.85 per 1M1000K ctx

Reserved

Ottili Spire 1.2

Reserved flagship of the Ottili AI Platform — not yet available.

Pricing at launch1000K ctx

Developer experience

Switch base URL. Keep your code.

Ottili AI Platform speaks the OpenAI request format. Any OpenAI-compatible client — Ottili Coder, OpenCode, Cline, Aider, the official SDKs — works by pointing its base URL at ai.ottili.one/api/v1.

Create API key Read the docs

curl https://ai.ottili.one/api/v1/chat/completions \
  -H "Authorization: Bearer $OTTILI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "ottili/vale-1.2",
    "messages": [{"role": "user", "content": "Write a TypeScript email validator."}]
  }'

Frequently asked questions

Can't find what you're looking for? Read the full documentation.

The shared AI model execution layer of Ottili ONE: one OpenAI-compatible API for Ottili's own models — Vale, Cairn and the Auto router — with usage tracking, API keys and prepaid billing.

Sign in, create an API key from the dashboard, add credits, and point any OpenAI-compatible client at https://ai.ottili.one/api/v1. No separate SDK is required.

Ottili Vale (efficient, everyday work) and Ottili Cairn (premium, deeper reasoning) are live today, plus Ottili Auto, which adaptively routes between them. Ottili Spire is reserved for a future flagship model and is not yet purchasable.

Yes — /v1/chat/completions and /v1/responses follow the OpenAI request and response shape, including streaming, structured outputs and tool calling, so existing OpenAI SDKs and clients work by changing only the base URL and key.

Ottili AI Platform is prepaid: you add credits, and each request's cost is reserved and deducted from your balance based on real per-model token pricing. There are no subscriptions or seat licenses, and a request is rejected up front if your balance can't cover it.

Yes — Ottili Coder's cloud coding surface already runs on Ottili AI Platform as its native model provider, using Vale, Cairn and Auto for real coding workloads.

Every request is scoped to an API key and resolved to a single company. Keys support per-key model and budget limits, and rate limiting is enforced per key and per company.

Documentation

Read the Ottili AI docs

Go deeper on each part of the Ottili AI surface. These guides explain how models, routing, missions, DWE specialists, tools, approvals, memory, credits, audit and cost control work in Ottili AI. Every link points to the public Ottili documentation; nothing here claims an availability the docs do not state.

Browse all Ottili AI docs

Less infrastructure.
More intelligence.

Integrate Ottili AI Platform with a few lines of code and run production workloads through one unified, OpenAI-compatible interface.

Start building Read the docs