66 lines
2.4 KiB
Markdown
66 lines
2.4 KiB
Markdown
|
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
|
|||
|
|
|