MiniRobotLanguage (MRL)
AIK.Ask Completion
Ask Claude AI via Completion Endpoint
Intention
The AIK.Ask Completion command seamlessly integrates Anthropic AI's Completion endpoint with SPR-Script, empowering scripts to leverage advanced AI for generating text completions.
This command provides access to a subset of Anthropic AI's models, offering enhanced capabilities within SPR-Script for a wide range of applications, from complex task analysis to rapid information retrieval.
Syntax
AIK.Ask Completion|P1[|P2]
AIK.Asc|P1[|P2]
Parameter Explanation
•P1 (Prompt): The text prompt for Anthropic AI's Completion endpoint. This prompt guides the AI in generating specific completions tailored to the script's needs.
•P2 (Return Variable or TOS): The destination for the AI-generated completion result. This can either be a specified variable or the Top Of Stack (TOS) if a variable is not provided.
Supported Models*
claude-2.1
•Description: Offers unparalleled capabilities for handling highly complex tasks with top-level performance, intelligence, fluency, and understanding.
•Strengths: Excels in providing comprehensive and detailed completions. Ideal for sophisticated queries and content generation.
•Multilingual: Yes
•Vision: Yes
•API Format: Messages & Text Completions API
•Comparative Latency: Slower than Claude 3 models but prioritizes detail in outputs.
•Context Window: 200K*
•Max Output: 4096 tokens
•Cost: Input / Output per MTok^ - $8.00 / $24.0
claude-2.0 (least censored)
•Description: Strikes a balance between intelligence and speed, crafted for enterprise workloads.
•Strengths: Delivers maximum utility at a cost-effective price point, suitable for scaled deployments.
•Multilingual: Yes
•Vision: Yes
•API Format: Messages & Text Completions API
•Comparative Latency: Offers a general-purpose speed, slightly slower than Claude 3 models.
•Context Window: 200K*
•Max Output: 4096 tokens
•Cost: Input / Output per MTok^ - $8.00 / $24.0
claude-instant-1.2 (most censored)
•Description: The fastest and most streamlined model, designed for near-instant responsiveness. Acts as a precursor to Claude Haiku.
•Strengths: Quick and precise, best for rapid responses and targeted performance.
•Multilingual: Yes
•Vision: No
•API Format: Messages & Text Completions API
•Comparative Latency: Fastest among the supported models, with a focus on promptness.
•Context Window: 100K**
•Max Output: 4096 tokens
•Cost: Input / Output per MTok^ - $0.80 / $2.40
* All Informations are dated 24.03.2024 from here:
Anthropic models Comparison
Example
'***********************************
' AIK.-Sample
'***********************************
AIK.Set Key|file
$$PRO+ Answer YES or NO, no explanation: Is 3+3=9?
AIK.Ask Completion|$$PRO|$$RET
DBP.$$RET
MBX.$$RET.
ENR.
Remarks
-
Limitations:
-
See also:
•