<< Click to Display Table of Contents >> Navigation:  3. Script Language > AI - Artificial Intelligence Commands > AIK. - Claude from Anthropic > AIK. - Message Endpoint > Template

---

MiniRobotLanguage (MRL)

 

AIK.Set Mes Model

Set the Model to use with the Messages Endpoint

 

 

Intention

 

Choose Model Parameter for Anthropic AI, Message Endpoint. In the Messages Endpoint, you can use all Models that are available from Anthropic for public use.

 

 

Syntax

 

 

AIK.Set Mes Model|P1

 

 

Parameter Explanation

 

P1 - Name of the Model that should be used. Use the "API Model Name" from the table below.

 

 

 

Supported Models*

 

Here's a table summarizing the available models and their properties within the Claude 3 family, as well as comparisons with legacy models:

 

 

API Model Name           |    Model                   |  Description                                                     

-------------------------|----------------------------|-----------------------------------------------------------------------

claude-3-opus-20240229   | Claude 3 Opus              | Most powerful model for highly complex tasks                          

claude-3-sonnet-20240229 | Claude 3 Sonnet            | Ideal balance of intelligence and speed for enterprise workloads      

claude-3-haiku-20240307  | Claude 3 Haiku             | Fastest and most compact model for near-instant responsiveness        

claude-2.1               | Claude 2.1                 | Updated version of Claude 2 with improved accuracy                    

claude-2.0               | Claude 2                   | Predecessor to Claude 3, offering strong all-round performance        

claude-instant-1.2       | Claude Instant 1.2         | Our cheapest small and fast model, a predecessor of Claude Haiku.     

 

 

More Details:

 

Model               | Description                                                        | Strengths                                                       | Multilingual | Vision | Latest API model name     | API format                         | Comparative latency                                         | Context window | Max output  | Cost (Input / Output per MTok^) | Training data cut-off

---------------------|-------------------------------------------------------------------|-----------------------------------------------------------------|--------------|--------|---------------------------|------------------------------------|-------------------------------------------------------------|----------------|-------------|---------------------------------|-----------------------

Claude 3 Opus        | Most powerful model for highly complex tasks                      | Top-level performance, intelligence, fluency, and understanding | Yes          | Yes    | claude-3-opus-20240229    | Messages API                       | Moderately fast                                             | 200K*          | 4096 tokens | $15.00 / $75.00                 | Aug 2023

Claude 3 Sonnet      | Ideal balance of intelligence and speed for enterprise workloads  | Maximum utility at a lower price, dependable.                   | Yes          | Yes    | claude-3-sonnet-20240229  | Messages API                       | Fast                                                        | 200K*          | 4096 tokens | $3.00 / $15.00                  | Aug 2023

Claude 3 Haiku       | Fastest and most compact model for near-instant responsiveness    | Quick and accurate targeted performance                         | Yes          | Yes    | claude-3-haiku-20240307   | Messages API                       | Fastest                                                     | 200K*          | 4096 tokens | $0.25 / $1.25                   | Aug 2023

Claude 2.1           | Updated version of Claude 2 with improved accuracy                | Legacy model - performs less well than Claude 3 models          | Yes          | No     | claude-2.1                | Messages & Text Completions API    | Slower than Claude 3 model of similar intelligence          | 200K*          | 4096 tokens | $8.00 / $24.0                   | Early 2023

Claude 2             | Predecessor to Claude 3, offering strong all-round performance    | Legacy model - performs less well than Claude 3 models          | Yes          | No     | claude-2.0                | Messages & Text Completions API    | Slower than Claude 3 model of similar intelligence          | 200K*          | 4096 tokens | $8.00 / $24.0                   | Early 2023

Claude Instant 1.2   | Our cheapest small and fast model, a predecessor of Claude Haiku. | Legacy model - performs less well than Claude 3 models          | Yes          | No     | claude-instant-1.2        | Messages & Text Completions API    | Slower than Claude 3 model of similar intelligence          | 200K*          | 4096 tokens | $0.80 / $2.40                   | Early 2023

 

Notes:

- *~150K words, ~680K unicode characters (indicated by "200K*").

- ^Costs are indicated per million tokens (MTok).

 

 

This comparison aims to guide you in selecting the right model based on your needs regarding complexity, latency, cost, and performance.

 

 

 

 

* All Informations are dated 24.03.2024 from here:
Anthropic models Comparison

 

 

 

 

Example

 

'***********************************

' AIK.-Sample

'***********************************
AIK.Set Key|file

AIK.SetTemperature|0.7

AIK.Set Mes Model|claude-2.0

$$PRO+ Please write a short Twitter Post as a rhyme.

$$PRO+ in german and add lots of emojis about the 

$$PRO+ great epic failures of Robert Habeck.

AIK.Ask Claude|$$PRO|$$RET

AIC.Get Raw Output|$$RAW

DBP.$$RET

MBX.$$RAW

DBP.$$RAW

MBX.$$RET

ENR.

 

 

 

 

 

Remarks

 

-

 

 

Limitations:

 

-

 

 

See also:

 

•