As of: October 7, 2026
Logicc brings leading AI models from different providers together on one platform. You can pick a model manually or use SmartSelectAI. SmartSelectAI automatically picks a suitable model for your request.
All models available in Logicc are hosted in the EU. Data related to their use is processed only within the EU.
Important: All Claude Opus models are only available from the Max plan.
Admins decide which models are available to the members of their organization:
https://app.logicc.com/organization-profile/models
There you control two levels:
Model providers: Here you set which providers your organization may use models through, for example Azure Foundry, Vertex AI, Amazon Bedrock, OpenAI or Mistral.
Models: Here you turn individual models on or off for users to choose in the chat.
A model can be available through several providers. The note under the model name shows how many providers can supply the model and how many of them are active for your organization. 2/2 Modellanbieter means, for example: The model is available through two providers, both are active. 1/2 Modellanbieter means: Only one of the two providers is active.
If you turn off a provider, your model settings stay saved. Affected models can only be used again once an active provider is available.
The overview helps you choose manually. If you are not sure, SmartSelectAI is usually the best choice. The model names match the spelling in the Logicc interface. As a rule, a higher version number means a newer version.
The newer GPT models come in different variants: Sol is meant for demanding tasks, Terra for balanced daily work and Luna for fast, low-cost routine tasks.
Model | Short profile | Typical uses |
|---|---|---|
GPT-6.1 Sol | Newest Sol version in Logicc for demanding tasks. | Complex analysis, tasks with several steps, high-quality write-ups and programming. |
GPT-6 Sol | Powerful GPT-6 model for complex tasks. According to OpenAI, it makes significantly fewer factual errors than GPT-5.6 Sol. | Complex analysis, data evaluation, tasks with several steps and programming. |
GPT-6 Luna | Fast, low-cost GPT-6 model for clearly defined tasks. | Summaries, pulling information out of texts, short answers and routine questions. |
GPT-5.6 Sol | Powerful GPT-5.6 variant for especially demanding tasks. | Complex analysis, tasks with several steps, high-quality write-ups and programming. |
GPT-5.6 Terra | Balanced GPT-5.6 variant for daily professional work. | Texts, summaries, structured analysis, planning and general office tasks. |
GPT-5.6 Luna | Low-cost GPT-5.6 variant for routine tasks. | Email drafts, short summaries, text variants and standard tasks. |
GPT-5.5 | Model for especially demanding tasks where quality matters more than speed. | Difficult document work, high-quality wording, deep analysis and complex planning. |
GPT-5.4 | Powerful GPT model for demanding all-round tasks. | Complex texts, professional write-ups, programming and detailed analysis. |
GPT-5.4 mini | Smaller GPT-5.4 variant with good quality for routine tasks. | Regular writing tasks, summaries, internal notes and short analyses. |
GPT-5.2 | Built for tasks with several reasoning steps. | More demanding analysis, decision support and structured problem solving. |
GPT-5.1 | GPT variant for general quality tasks. | Precise drafts, structured answers, longer texts and tasks with clear requirements. |
GPT-5 | Broad all-round model for language, structure and analysis. | Fully written texts, summaries, collecting ideas, planning and explanations. |
GPT-5 mini | Fast all-round model for everyday tasks. | Email drafts, chat replies, short analyses, text variants and standard summaries. |
o3 | Reasoning model for tasks with many intermediate steps. | Logic, math, technical analysis, traceable derivations and complex decisions. |
Claude models come in three tiers: Haiku is fast and low-cost, Sonnet is the balanced all-round model and Opus is the most thorough tier for the most demanding tasks. Claude models are especially strong at careful text and document work.
All Opus models are only available from the Max plan.
Model | Short profile | Typical uses |
|---|---|---|
Claude Opus 5.5 | Newest and strongest Claude model in Logicc. From the Max plan. | Tasks that need a lot of care and judgment: deep document analysis, complex technical texts and difficult trade-offs. |
Claude 5 Opus | Opus model of the Claude 5 generation for especially complex tasks. From the Max plan. | Long documents, strategic write-ups, demanding reviews and complex trade-offs. |
Claude 4.8 Opus | Powerful Opus variant for very demanding analysis and writing tasks. From the Max plan. | Deep document analysis, complex professional write-ups and large amounts of text. |
Claude 4.7 Opus | Opus model for especially complex analysis and writing tasks. From the Max plan. | Long documents, strategic write-ups, high-quality summaries and demanding reviews. |
Claude 4.6 Opus | Opus variant for very careful analysis. From the Max plan. | Extensive materials, complex trade-offs, precise structuring and review tasks. |
Claude 4.5 Opus | Strong Opus model for deep analysis and complex documents. From the Max plan. | Contract analysis, chains of argument, risk overviews and demanding technical texts. |
Claude Sonnet 5.5 | Newest Sonnet version in Logicc. According to Anthropic, the best combination of speed and intelligence, and faster than Claude 5 Sonnet. | Everyday professional tasks, creating and revising documents, structured evaluations and text work. |
Claude 5 Sonnet | Sonnet model of the Claude 5 generation for demanding professional tasks. | Long documents, structured reviews, careful write-ups and technical texts. |
Claude 4.6 Sonnet | Sonnet variant for more demanding standard tasks. | Longer documents, cleanly organized answers, analysis with requirements and text revision. |
Claude 4.5 Sonnet | Balanced model for precise text and document work. | Document analysis, structured summaries, email drafts and technical texts. |
Claude 4.5 Haiku | Fast, low-cost Claude model for light and medium tasks. | Short document summaries, quick text work, simple classification and routine questions. |
Flash models are fast and can process a lot of text at once, for example long documents or several files at the same time. They also understand images. Pro is the more thorough tier for more complex tasks.
Model | Short profile | Typical uses |
|---|---|---|
Gemini 2.5 Pro | More thorough Gemini model for complex tasks with a lot of material. | Large case files, long documents, combined image and text analysis and demanding evaluations. |
Gemini 3.8 Flash | Newest and most powerful Flash version in Logicc. Especially strong at tasks with several steps and many documents. Works more thoroughly than 3.7 Flash and uses a bit more budget for it. | Extensive document work, multi-step evaluations and large collections of material. |
Gemini 3.7 Flash | Flash version with good document understanding. | Long documents, quick evaluations and image and text tasks. |
Gemini 3.6 Flash | Efficient Flash version for fast processing. | Long inputs, image and text tasks and quick analysis of larger materials. |
Gemini 3.5 Flash | Flash version for fast processing with a lot of material. | Extensive documents, quick image and text evaluations. |
Gemini 2.5 Flash | Fast model for a lot of material and image and text tasks. | Long inputs, image and text tasks and quick analysis of larger materials. |
Mistral is a European AI provider from France. The models work well for efficient text work and multilingual content.
Model | Short profile | Typical uses |
|---|---|---|
Mistral Medium 3.5 | Medium variant for more demanding standard tasks. | More thorough analysis, better drafts, longer texts, structured evaluations and multilingual content. |
Mistral Small 4 | Small model for efficient everyday work. | Recurring text tasks, translations, internal notes and simple analysis. |
If you don't want to decide manually, use SmartSelectAI. Logicc then automatically picks a suitable model for your request.
For fast standard tasks: GPT-6 Luna, GPT-5.6 Luna, GPT-5 mini, GPT-5.4 mini, Claude 4.5 Haiku or Mistral Small 4.
For balanced all-round results: Claude Sonnet 5.5, GPT-5.6 Terra, GPT-5.4, Claude 5 Sonnet or Mistral Medium 3.5.
For especially demanding tasks: GPT-6.1 Sol, GPT-6 Sol, GPT-5.6 Sol, GPT-5.5 or Claude Opus (from the Max plan).
For a lot of material or images: Gemini 3.8 Flash, another Gemini Flash version or Gemini 2.5 Pro.
For tasks with many reasoning steps: o3 or one of the Sol models.
You can find more guidance in the article Choosing the right model.
If your organization has signed an agreement under § 203 StGB (Germany), § 121 StGB (Austria) or Art. 321 StGB (Switzerland), only the models whose providers are included in that agreement are available. Most models stay available. More on this in the article Agreements.
If you don't see a certain model, it may be turned off for your organization, no active provider is available for it, or it is not included in your plan. In that case, contact an admin of your organization or [email protected].
