2.4 KiB
2.4 KiB
PricingModel
Pricing for Kaizen agents
KaizenAgentic Pricing Model
1. Base Assumption
- Token Cost (C): The unit price per token for the underlying foundation model (e.g., OpenAI GPT-4o, Anthropic Claude, etc.).
- KaizenAgentic charges are always calculated as a multiple of this base token cost.
2. Capability Multipliers
Each subagent is classified by its capability tier, which reflects complexity, optimization overhead, and real-world utility.
| Tier | Agent Capability | Multiplier (x) | Example Use Case |
|---|---|---|---|
| 1x | Baseline wrapper agent | 1× | Simple automation around base LLM calls |
| 2x | Enhanced agent | 2× | Adds logging, minimal optimization, lightweight feedback loops |
| 3x | Professional agent | 3× | Integrated metrics, test coverage deltas, developer UX signals |
| 4x | Expert agent | 4× | Adaptive refinement, A/B testing, rollback mechanisms |
| 5x | KaizenAgent premium | 5× | Full meta-optimization loop, cross-subagent orchestration |
3. Pricing Formula
\text{KaizenAgentic Price per Token} = C \times M
Where:
- C = cost per token of the underlying LLM
- M = capability multiplier (2x–5x)
Example:
- GPT-4o base token = $0.01 / 1K tokens
- KaizenAgent Premium (5x) = $0.05 / 1K tokens
4. Service Tiers
On top of token-based billing, KaizenAgentic can introduce subscription layers to cover operational support:
- Free Tier → 1x baseline agents, capped usage, no optimization feedback.
- Pro Tier → 2x–3x agents, includes monitoring dashboards.
- Enterprise Tier → 4x–5x agents, includes dedicated KaizenAgent meta-optimization + SLAs.
5. Value Rationale
- Fair: Always anchored in base token price (transparent to clients).
- Scalable: Higher capability → higher multiplier → more value.
- Predictable: Clients can forecast spend by capability tier, independent of vendor-specific LLM pricing changes.
- Flexible: Basemodel transparent to avoid basemodel lockin supporting various providers (ChatGPT, Claude, Cursor, etc.).
xxx