Built for teams that take
engineering seriously.
Choose the deployment mode that fits your security, compliance, and operational needs.
Built for teams that take
Most Popular
Cloud
Managed by us. Loved by you.
- Fully hosted SaaS — zero infrastructure work
- Automatic updates and improvements
- Managed scaling and monitoring
- EU data residency for GDPR
- All platform features included
- Priority support
Enterprise
Self-Hosted
Your infrastructure. Your rules.
- Deploy on your own cloud or on-premise
- Full data sovereignty — nothing leaves your network
- Air-gap / offline deployment supported
- Custom model providers and integrations
- No per-seat charges
- Dedicated support and onboarding
Feature comparison
| Feature | Cloud | Self-Hosted |
|---|---|---|
| Job Pipeline | ||
| Expert Agents | ||
| Guardrails | ||
| Dashboard | ||
| Knowledge Management | ||
| Multi-Model Support | ||
| Integrations | ||
| Benchmark Lab | ||
| CLI & API | ||
| Offline Mode | ||
| Full Data Sovereignty | Isolated | |
| Managed Updates | Manual | |
| SLA | 99.9% | Self-managed |
Total cost of ownership
Where the money and time actually go — beyond the subscription price.
| — | Cloud | Self-Hosted |
|---|---|---|
| Capex | None | Infra provisioning, initial setup time |
| Opex | Subscription + model tokens | Infra + ops + model tokens |
| Ops burden | We run it | Your team runs it — upgrades, backups, monitoring |
| Model choice | Curated providers, swap per project | Any provider, including local / offline |
| Data residency | EU region, isolated per customer | Wherever you deploy — on-prem, private cloud, air-gap |
| SLA | 99.9% uptime | Self-managed |
What a job actually costs
Representative per-job costs from real workloads — the model tokens dominate, not the subscription.
| Job type | Typical cost | What drives it |
|---|---|---|
| Security audit | $0.20 – $1.00 | One analysis expert reads the repo, runs tools, submits findings. Read-heavy; no writes. |
| Feature job (median) | $2 – $8 | Decomposition + 3–6 mini-jobs running in parallel. Write-heavy; multi-turn coding loops. |
| Chat session | ≈ $0.05 | Single Haiku turn per message; full context cached after turn 1. |
Figures are indicative ranges from observed job medians on Claude Haiku 4.5 (fast tier) and Sonnet 4.6 (default tier). Deep-tier jobs on Opus or GPT-class frontier models run 3–5× higher. Check `/api/jobs/cost-estimate` for live medians.
Per-tier model pricing
Orqista routes each job to one of three tiers. Prices are passed through at cost — the dashboard shows live per-tier rates from your connected model providers.
| Tier | Typical model class | Input (per 1M tokens) | Output (per 1M tokens) |
|---|---|---|---|
| fast | Haiku-class / small local | $0.25 – $1.00 | $1.00 – $5.00 |
| default | Sonnet-class | $1.00 – $3.00 | $5.00 – $15.00 |
| deep | Opus / frontier | $5.00 – $15.00 | $25.00 – $75.00 |
Ranges below are the public per-1M-token rates across typical providers (Anthropic, Bedrock, Ollama-free). Local models on the fast tier drop the effective rate to zero at the cost of hardware.
How Orqista compares
Feature comparison with popular AI coding tools.
Orqista 9/10
Devin 3/10
Cursor 1/10
Copilot WS 1/10
SWE-agent 3/10
| Feature | Orqista | Devin | Cursor | Copilot WS | SWE-agent |
|---|---|---|---|---|---|
| Self-hosted / air-gap | |||||
| DAG execution (staged parallelism) | |||||
| Configurable guardrails | |||||
| Multi-model (swap providers) | |||||
| Custom workflow patterns | |||||
| GitLab + Jira integration | |||||
| Knowledge system | |||||
| Human-in-the-loop approval | |||||
| Cost tracking & budgets | |||||
| Open source |
Comparison based on publicly available information as of March 2026. Features may have changed.
Pricing FAQ
What's included in the Cloud plan?
Full platform access — all expert agents, integrations, guardrails, knowledge management, and the dashboard. We handle infrastructure, updates, and scaling.
Can I switch between Cloud and Self-Hosted?
Yes. Your workflows, expert agents, guardrails, and knowledge migrate seamlessly between deployment modes. No vendor lock-in by design.
Do you offer a free trial?
Contact us for trial access. We'll set you up with the right deployment mode for your team and walk you through the platform.
What AI models can I use?
Any model provider. Cloud deployments use managed AI services with EU data residency. Self-hosted deployments can use any provider, including local models for offline operation.
Is there a per-seat charge?
Self-hosted: no per-seat charges. Cloud: pricing is based on usage, not seats. Contact us for details that fit your team size.
Let's find the right fit
Tell us about your team and we'll recommend the best deployment option.
Request Early Access