Start free. Scale without surprises.
Every plan includes AI chat, web research, and automation tools, for AI and SI (superintelligence) workloads alike. On-site deployment is available for regulated enterprises — pay only for energy, not per token.
Pick your tier
All plans run on SLM-routed infrastructure — you get frontier-quality output at a fraction of the token cost. Upgrade or cancel anytime.
Free
- 500K free credits/day
- 2 widgets
- 10 shared data rows/day
Estimated savings: about 80% on the same work versus a frontier-only model.
Usage extension: every Token Saver run reports how much further it stretched your usage window.
Savings figure from one production account over 30 days, measured and reported, not projected. Your own runs show on /savings.
Starter
- 100K credits/month
- 5 widgets
- Web crawl up to 3 pages
- 50 shared data rows/day
Estimated savings: about 80% on the same work versus a frontier-only model.
Usage extension: every Token Saver run reports how much further it stretched your usage window.
Savings figure from one production account over 30 days, measured and reported, not projected. Your own runs show on /savings.
Standard
- 1M credits/month
- 20 widgets
- Deep data collection (up to 100 items)
- Web crawl up to 20 pages
- 250 shared data rows/day
Estimated savings: about 80% on the same work versus a frontier-only model.
Usage extension: every Token Saver run reports how much further it stretched your usage window.
Savings figure from one production account over 30 days, measured and reported, not projected. Your own runs show on /savings.
Pro
- 10M credits/month
- Unlimited widgets
- Deep data collection (up to 10,000 items)
- Unlimited web crawl pages
- Public / embeddable widgets
- 2,500 shared data rows/day
Estimated savings: about 80% on the same work versus a frontier-only model.
Usage extension: every Token Saver run reports how much further it stretched your usage window.
Savings figure from one production account over 30 days, measured and reported, not projected. Your own runs show on /savings.
Organization
- Unlimited widgets
- Unlimited deep collection
- Unlimited web crawl
- Public widgets
- 10,000 shared data rows/day
- Team management & roles
- Per-seat credit allocation, plus an optional shared org pool
Estimated savings: about 80% on the same work versus a frontier-only model.
Usage extension: every Token Saver run reports how much further it stretched your usage window.
Savings figure from one production account over 30 days, measured and reported, not projected. Your own runs show on /savings.
| Feature | Free | Starter | Standard | Pro | Organization |
|---|---|---|---|---|---|
| Widgets | 2 | 5 | 20 | Unlimited | Unlimited |
| Deep Collection | — | — | ✓ (100 items) | ✓ (10K items) | ✓ Unlimited |
| Web Crawl Pages | — | 3 | 20 | Unlimited | Unlimited |
| Public Widgets | — | — | — | ✓ | ✓ |
| Shared Data Rows/Day | 10 | 50 | 250 | 2,500 | 10,000 |
| Team Management | — | — | — | — | ✓ |
| Monthly Credits | 500K/day free | 100K | 1M | 10M | Per seat + pool |
| Token Saver (MCP) | |||||
| Loading live Token Saver terms… | |||||
| Live figures for your own account are in the entitlement block of every Token Saver MCP response, and on your billing page. Full Token Saver terms | |||||
Credits are cost-weighted — instant/fast models stretch a credit ~10× further than deep/frontier.
You don't pay for one big model on every task
Our router picks the right-sized model for each request — often a smaller, cheaper, or self-hosted model — instead of sending everything to one frontier model. Across the whole platform, here's the estimated difference.
⚡ See your own savings — try the live demo →Estimating…
Regulatory Compliance
Dedicated clusters include built-in compliance controls for regulated industries. Your infrastructure is designed to meet the following regulatory frameworks.
Upgrade any organization to a Dedicated Cluster for isolated compute, BYOK, tenant scoping, and full regulatory compliance controls.
Annual on-prem license by seat block. Includes full Docker bundle download and a signed activation code for onsite deployment enforcement.
- 5 licensed users per block ($240/user/yr)
- Full Docker bundle after Stripe checkout
- Signed activation code for onsite enforcement
Own your infrastructure. Own your costs.
Every major AI vendor is moving toward consumption-based billing just as enterprises are discovering their consumption is uncontrollable. Elis AI on-site deployment eliminates the token meter — you pay energy, not invoices. The table below includes us alongside competitors so you can make an honest comparison.
| Provider | On-site | Deployment | Per seat | 100 users/mo | Annual cost | Setup fee | Contract |
|---|---|---|---|---|---|---|---|
| Microsoft Copilot | Cloud only | Cloud SaaS only | $30/user/mo | $3,000/mo | $36,000/yr | None | Monthly |
| IBM Watsonx | Yes | Cloud or on-premise | Custom contract | $4,000+/mo est. | $50,000+/yr | $20,000–$100,000 | Annual |
| Harvey AI | Cloud only | Cloud SaaS only | $100+/user/mo | $10,000+/mo | $120,000+/yr | None | Annual |
| Palantir AIP | Yes | Cloud or on-premise | Enterprise contract | $40,000+/mo est. | $500,000+/yr | $100,000+ | Multi-year |
| Nightfall | Cloud only | Cloud SaaS only | $20+/user/mo | $2,000+/mo | $24,000+/yr | None | Annual |
| Elis AI (Cloud) | Cloud only | Cloud SaaS | From $8/user/mo | From $800/mo | From $9,600/yr | None | Monthly |
| Elis AI (On-site) | Yes | Your servers, your data | Energy cost only | ~$200–800/mo energy* | ~$2,400–9,600/yr energy* | One-time setup fee | None — you own it |
* On-site energy estimate based on typical GPU server draw (~500W) at $0.10/kWh for an 8-GPU node running 24/7. Actual costs vary by hardware and energy rates. Competitor figures are public estimates as of 2026 and may not reflect negotiated enterprise rates.
For regulated enterprises with their own infrastructure, on-site deployment eliminates the token meter. No per-call charges, no usage overages, no surprise invoices. The more your team uses it, the better the ROI — instead of the worse.
Dedicated cloud gives you a private Elis environment on our infrastructure — isolated compute, custom domain, SSO, and SLA guarantees — without the capital cost of on-site hardware.
Need ITAR / EAR, air-gapped, or custom clearance deployment?
We work with regulated teams in defense, healthcare, and legal on bespoke deployment architectures. These engagements are scoped individually — contact us to start the conversation.