Pricing
From 95% to 99.5%. The Last Mile of Agent Reliability.
From 85.7% to 100% on hard code. +20.8pp on reasoning tasks. 12.2x deeper chains that don't decay. 311 cognitive abilities behind one endpoint.
Haki
Most Populardominant will · multi ability
Compound reasoning scaffolds for production agents.
- ✓All 6 reasoning dimensions. 311 cognitive abilities
- ✓4 synergized abilities per call with compound suppression
- ✓Synergy chains: primary + dependency + amplifier + alternative
- ✓10,000 calls/month
- ✓100 requests/minute
- ✓Full scaffolds: failure gates, procedures, checkpoints, vectors
What it solves
Real problems don't fit one dimension. A supply chain decision is a causal problem, a timing problem, and a constraint problem at the same time. Haki gives your agent four perspectives in one call: a primary scaffold sets the reasoning direction, a dependency grounds it in prerequisites, an amplifier deepens the analysis, and an alternative challenges the conclusion before it solidifies. Your agent doesn't just think. It thinks from four angles and arrives at one answer.
Ki
focused energy · single ability
One reasoning scaffold per call. Covers most agent failures.
- ✓All 6 reasoning dimensions. 311 abilities
- ✓Full cognitive scaffold per call: procedure, failure gate, checkpoint
- ✓Suppression + amplification vectors. Reasoning topology
- ✓5,000 calls/month
- ✓100 requests/minute
What it solves
Your agent sends its task description. The API returns the most relevant cognitive ability: a negative gate that names the failure pattern to avoid, a step-by-step reasoning procedure, a falsification checkpoint to verify the output, and suppression vectors that block the exact reasoning shortcuts your task is vulnerable to.
Enterprise
Dedicated cognitive infrastructure for agent fleets.
- ✓All 6 dimensions. Unlimited abilities per call
- ✓Unlimited volume
- ✓Private deployment. Dedicated instance
- ✓Custom ability engineering for your domain
- ✓SLA + dedicated support
What it solves
At scale, reasoning degradation compounds across your fleet. Enterprise brings the scaffolding layer inside your perimeter. Abilities tuned to your specific failure modes, private inference, and dedicated engineering to ensure every agent in your fleet maintains reasoning fidelity under your production constraints.
Start with 100 free calls. No card required. Create a free account →
Your agent doesn't lack knowledge. It loses reasoning quality over time.
Every LLM degrades its reasoning as tasks get longer. Attention drifts, failure patterns go unchecked, conclusions get accepted without verification. The Logic API is a runtime reasoning layer your agent calls to maintain cognitive fidelity. Each ability is an engineered scaffold: a negative gate that names the failure to avoid, a procedure that structures execution, a checkpoint that verifies the output, and suppression vectors that block the reasoning shortcuts your task is vulnerable to.
Measured across 10 professional domains: +7.4 correctness lift, +158% self-monitoring, +49% verification behavior. On hard competitive programming: 85.7% to 100% pass rate, zero regressions. Starting at €19/month.
Frequently Asked
More answers in the full FAQ · Quickstart guide · Use cases · Try it in the Playground
Ki or Haki. Which do I need?
Start with Ki. One ability per call handles most reasoning failures. The API identifies the highest-scoring cognitive scaffold for your task and returns it complete with suppression vectors, amplification signals, reasoning procedure, and falsification checkpoint. Upgrade to Haki when your tasks span multiple reasoning dimensions simultaneously. Haki returns 4 synergized abilities that challenge and reinforce each other, preventing the tunnel vision that single-ability mode can produce.
What happens when I hit my call limit?
You receive a 429 response with a clear message and a link to upgrade. Free tier: 100 calls total. Ki: 5,000 calls/month, resets on billing cycle. Haki: 10,000 calls/month. Burst rate: 100 requests per minute across all tiers. Upgrade anytime from your dashboard.
Is there a free tier?
Yes. 100 API calls, single mode, no credit card required. Create an account, generate a key from your dashboard, and start calling the API immediately. Same endpoint, same abilities, same response format as paid plans.
What uptime do you guarantee?
We target 99.9% availability on a global edge network with automatic failover. If the API is unreachable, your agent continues on native LLM reasoning. Ejentum is a reasoning enhancement layer, not a hard dependency in your execution chain.
How do I measure the impact?
Run 50 to 100 representative tasks before and after integrating. Compare task completion rate, reasoning chain quality (does the agent self-monitor, verify, consider alternatives?), and token cost per task. Our benchmarks show +7.4 correctness lift, +158% self-monitoring, and +49% verification behavior across 10 professional domains.
Does it work with any LLM?
The abilities are model-agnostic. They are structured reasoning scaffolds, not model-specific prompts. The cognitive operations (suppression, amplification, falsification checkpoints) work with any instruction-following model. We test across major model families.
What exactly does the API return?
A JSON object containing a pre-rendered cognitive scaffold. Each ability includes: a negative gate (the failure pattern to avoid), a step-by-step reasoning procedure, a reasoning topology, a target pattern (what correct reasoning looks like), suppression and amplification vectors, and a falsification test. You inject this into your agent's context before task execution.