Practice, 1v1, AI builder, puzzle rush, AI tutor.
Classrooms for your CS course, free during pilot.
Evidence-linked assessments; public access is waitlist-only.
Pay per question.
5 candidates free.
We are still finishing the product and onboarding assessment teams through the waitlist. At launch, every recruiter account will include five free candidates, followed by the pricing structure below across No-AI, Rena chat, Agentic, and Agentic+ questions.
Free
Five free candidates when self-serve access launches.
- 1 workspace per account
- 5 candidates lifetime, no expiry
- Mix No-AI / Rena / Agentic / Agentic+ freely
- Up to 2 invites per 24h
- No card required
Starter
$150 of credit each month.
- $150 monthly credit pool
- Multiple workspaces
- Mix modes freely up to your pool
- Overage at rack rates ($5 / $10 / $30 / $60)
- No long-term contract
Scale
$700 of credit + −10% at 100+ questions/mo.
- $700 monthly credit pool
- Multiple workspaces
- −10% volume discount auto-applies at 100+ questions/mo
- Per-question AI-mode selection
- Spend cap + email alerts
Pro
$3,000 of credit + −10% baked in.
- $3,000 monthly credit pool
- Multiple workspaces
- −10% baked into every overage attempt
- −25% at 1,000+ questions/mo
- Priority support
Join your company team
Recruiting teammates join from the personal, email-bound link sent by a workspace owner or admin.
A personal link is required
For security, each invitation is bound to one verified email address, expires, and can be accepted once.
Candidate assessment invitations are separate and open directly from their emailed links.
Have a coupon code?
Pilot or partner assessment-credits code from AlgoArena, not a teammate workspace invitation.
Four AI modes. Four price points.
Modes are set per question, not per assessment, so one OA can mix all four. These are the intended launch rates; no public assessment checkout is open yet.
Candidate codes solo. Strong, deliberate signal: locks every AI surface. Best for high-volume top-of-funnel screening.
Text chat partner that talks through tradeoffs but never writes code for the candidate. Tests reasoning + prompt-writing.
Full agent partner. Candidate prompts; agent scaffolds and edits. Measures the realistic senior-eng workflow.
The strongest models with double the AI runway. Frontier lineup (Claude Opus 4.8 + Sonnet 5) for roles where frontier-model fluency is the signal.
Planned rack rates shown. Paid tiers will include monthly credit and automatic volume discounts.
Run the numbers.
Model your expected monthly mix against the planned launch structure while the product remains waitlist-only.
Preview planned monthly cost
| Plan | Subscription | Overage | Total |
|---|---|---|---|
Starter $150 credit included | $99 | $450.00 | $549.00 $7.32/question avg |
ScaleLowest estimate $700 credit included | $399 | — | $399.00 $5.32/question avg |
Pro $3000 credit included | $1499 | — | $1499.00 $19.99/question avg |
Launch estimate only. Volume discounts are planned to apply automatically beyond each monthly threshold, with overage billed at the same per-question rate rather than a penalty rate.
What the price follows.
The launch model prices the assessment condition actually assigned to each question. It does not turn unvalidated market statistics into product proof.
- No-AI prices the editor, tests, browser, and integrity workflow without an AI assistant.
- Rena-guided adds bounded conceptual guidance while keeping direct code generation outside the condition.
- Agentic prices the model, tool, terminal, search, delegation, replay, and evidence workflow needed for AI-enabled work.
- Agentic+ prices the same agentic workflow on the frontier model lineup with double the AI runway.
- Recruiters can mix conditions by question. Planned volume discounts will apply at launch; current access is through the assessment waitlist.
Common questions
What if I run out of free candidates?▾
At launch, you will be able to upgrade to any paid tier. The planned free quota is 5 candidates total for the lifetime of your account, not a monthly or yearly allowance. Each invited candidate will count as one, regardless of question count.
How does overage work?▾
At launch, each paid tier will include a monthly credit pool. Additional attempts will use the same rack rate (Starter / Scale) or the tier's baked-in discount (Pro), with volume discounts applied automatically.
Can I cap my monthly spend?▾
The launch billing design includes a hard monthly cap, pre-invite cost estimates, and email alerts at 50%, 80%, and 100% of the cap. Waitlist teams can help us validate those controls before general availability.
How are AI modes priced per question?▾
The planned model prices each question by its AI condition. For example, 3× No-AI ($15) + 2× Agentic ($60) + 1× Agentic+ ($60) would total $135 before plan or volume discounts.
What happens to data when an attempt fails or the candidate doesn't finish?▾
The launch policy is to bill only completed attempts. Half-finished, abandoned, or rejected attempts will not count.