Intro
Domo offers multiple versions of AI models, each with distinct capabilities and optimizations. This guide helps you understand what changed between versions and choose the right model for your needs.
Model families covered:
- Anthropic Claude Sonnet (versions: 3.7, 4, 4.6)
- DomoGPT—Large (version: 2.2)
- DomoGPT—Medium (versions: v1, v1.1, v1.2, v2, v2.1, v2.2)
- DomoGPT—Small (versions: v1, v1.1, v2.1)
- GPT (versions: 3.5-Turbo, 4o, 4o-mini)
- Gemini (versions: 2.5-Flash, 2.5-Pro, 3-Pro)
- Llama (versions: 3.1-405B, 4-Maverick)
Understand Anthropic Claude Sonnet
Available versions: 3.7, 4, 4.6
Claude Sonnet models by Anthropic offer an optimal balance between intelligence and speed. They excel at complex reasoning, coding, content generation, and nuanced analysis. Sonnet is ideal for most business applications requiring sophisticated AI capabilities.
Compare Versions
| Version | Max response length | Context window | Supports images/docs | Use case |
|---|
| 3.7 | 64,000 tokens (~48,000 words or 128 pages) | 200,000 | Yes | Long-form content, document analysis |
| 4 | 16,384 tokens (~12,288 words) | 200,000 | Yes | General tasks |
| 4.6 | 64,000 tokens (~48,000 words or 128 pages) | 1,000,000 | Yes | Coding, computer use, long-context reasoning, agent planning, knowledge work, and design |
Review Version Changes
Upgrade from 3.7 to 4
- Response length reduced — maximum output decreased from 64,000 tokens (~48,000 words or 128 pages) to 16,384 tokens (~12,288 words)
Upgrade from 4 to 4.6
- Longer responses — maximum output increased from 16,384 tokens (~12,288 words) to 64,000 tokens (~48,000 words or 128 pages)
- Larger context window — increased from 200,000 to 1,000,000 tokens, allowing you to process much longer documents or maintain longer conversation histories
- Expanded capabilities — improvements across coding, computer use, long-context reasoning, agent planning, knowledge work, and design
Choose a Version
Latest version: Claude Sonnet 4.6—Recommended for most use cases, especially when you need long-context reasoning or comprehensive responses.
Understand DomoGPT—Medium
Available versions: v1, v1.1, v1.2, v2, v2.1, v2.2
DomoGPT Medium models are powered by Anthropic Claude and offer a balance of capability and performance. They’re ideal for complex reasoning tasks, content generation, and data analysis.
Compare Versions
| Version | Claude model | Max response length | Context window | Supports images/docs | Use case |
|---|
| v1 | Sonnet 3 | 4,096 tokens | 200,000 | Yes | Document analysis |
| v1.1 | Sonnet 3.5 | 8,192 tokens (~6,144 words) | 200,000 | Yes | Document analysis |
| v1.2 | Sonnet 3.7 | 64,000 tokens (~48,000 words or 128 pages) | 200,000 | Yes | Long-form content, document analysis |
| v2 | Sonnet 4 | 64,000 tokens (~48,000 words or 128 pages) | 200,000 | Yes | Long-form content, document analysis |
| v2.1 | Sonnet 4.5 | 64,000 tokens (~48,000 words or 128 pages) | 200,000 | Yes | Long-form content, document analysis |
| v2.2 | Sonnet 4.6 | 64,000 tokens (~48,000 words or 128 pages) | 1,000,000 | Yes | Long-context reasoning, agent planning, knowledge work, and design |
Review Version Changes
Upgrade from v1 to v1.1
- Longer responses — maximum response length increased from 4,096 tokens to 8,192 tokens (~6,144 words); useful for more detailed answers and longer content generation
Upgrade from v1.1 to v1.2
- Longer responses — maximum response length increased from 8,192 tokens (~6,144 words) to 64,000 tokens (~48,000 words or 128 pages); useful for more detailed answers and longer content generation
Upgrade from v1.2 to v2
- Minor updates — this version includes minor technical updates; core capabilities remain similar to the previous version
Upgrade from v2 to v2.1
- Minor updates — this version includes minor technical updates; core capabilities remain similar to the previous version
Upgrade from v2.1 to v2.2
- Larger context window — increased from 200,000 to 1,000,000 tokens, allowing you to process much longer documents or maintain longer conversation histories
- Expanded capabilities — improvements across long-context reasoning, agent planning, knowledge work, and design
Choose a Version
Latest version: v2.2—Recommended for most use cases, especially when you need comprehensive responses or detailed content generation.
Understand DomoGPT—Small
Available versions: v1, v1.1, v2.1
DomoGPT Small models are powered by Anthropic Claude and optimized for speed and cost-efficiency. They’re ideal for simpler tasks, quick responses, and high-volume processing.
Compare Versions
| Version | Claude model | Max response length | Context window | Supports images/docs | Use case |
|---|
| v1 | Sonnet 3 | 4,096 tokens | 200,000 | Yes | Document analysis, batch processing |
| v1.1 | Sonnet 3.5 | 8,192 tokens (~6,144 words) | 200,000 | Text only | General tasks |
| v2.1 | Sonnet 4.5 | 64,000 tokens (~48,000 words or 128 pages) | 200,000 | Yes | Long-form content, document analysis |
Review Version Changes
Upgrade from v1 to v1.1
- Longer responses — maximum response length increased from 4,096 tokens to 8,192 tokens (~6,144 words); useful for more detailed answers and longer content generation
- Input capabilities reduced — the model no longer accepts documents (PDF, Word, and so on) or images (JPG, PNG, and so on)
- Batch processing removed — this model can no longer be used in Domo dataflows for batch processing; use the real-time API only
If you rely on document/image input or batch processing, consider staying on v1 or using a different model family.
- Longer responses — maximum response length increased from 8,192 tokens (~6,144 words) to 64,000 tokens (~48,000 words or 128 pages)
- Image and document support restored — the model once again accepts documents (PDF, Word, and so on) and images (JPG, PNG, and so on)
Choose a Version
Latest version: v2.1—Recommended for most use cases, especially when you need comprehensive responses or detailed content generation.
Understand GPT
Available versions: 3.5-Turbo, 4o, 4o-mini
OpenAI’s GPT models are industry-leading language models. The family includes various versions optimized for different use cases—from the cost-effective GPT-3.5-Turbo for high-volume tasks to GPT-4o for advanced multimodal capabilities.
Compare Versions
| Version | Max response length | Context window | Supports images/docs | Use case |
|---|
| 3.5-Turbo | 8,196 tokens (~6,147 words) | 8,196 | Text only | General tasks |
| 4o | 16,384 tokens (~12,288 words) | 128,000 | Yes | Document analysis |
| 4o-mini | 16,384 tokens (~12,288 words) | 128,000 | Yes | Document analysis |
Review Version Changes
Upgrade from 3.5-Turbo to 4o
- Longer responses — output limit increased from 8,196 tokens (~6,147 words) to 16,384 tokens (~12,288 words)
- New input support — the model now accepts documents (PDF, Word, and so on) and images (JPG, PNG, and so on)
- Larger context window — increased from 8,196 to 128,000 tokens, allowing you to process longer documents or maintain longer conversation histories
Upgrade from 4o to 4o-mini
- Minor updates — this version includes minor technical updates; core capabilities remain similar to the previous version
Choose a Version
Latest version: 4o-mini—Suitable for general-purpose tasks with moderate response length requirements.
Understand Gemini
Available versions: 2.5-Flash, 2.5-Pro, 3-Pro
Google’s Gemini models are next-generation multimodal AI models. They offer impressive context windows and come in different variants: Flash for speed and efficiency, and Pro for maximum capability and reasoning power.
Compare Versions
| Version | Max response length | Context window | Supports images/docs | Use case |
|---|
| 2.5-Flash | 65,536 tokens (~49,152 words or 131 pages) | 1,048,576 | Yes | Long-form content, document analysis |
| 2.5-Pro | 65,536 tokens (~49,152 words or 131 pages) | 1,048,576 | Yes | Long-form content, document analysis |
| 3-Pro | 65,536 tokens (~49,152 words or 131 pages) | 1,048,576 | Yes | High-reasoning tasks and advanced coding |
Review Version Changes
Upgrade from 2.5-Flash to 2.5-Pro
- Minor updates — this version includes minor technical updates; core capabilities remain similar to the previous version
Upgrade from 2.5-Pro to 3-Pro
- Enhanced reasoning and coding — optimized for high-reasoning tasks and advanced coding use cases
Choose a Version
Latest version: 3-Pro—Recommended for most use cases, especially when you need comprehensive responses or detailed content generation.
Understand Llama
Available versions: 3.1-405B, 4-Maverick
Meta’s Llama (Large Language Model Meta AI) models are open-weight models that offer strong performance across various tasks. They’re particularly useful for organizations that need transparency and control over their AI infrastructure.
Compare Versions
| Version | Max response length | Context window | Supports images/docs | Use case |
|---|
| 3.1-405B | 2,048 tokens | 128,000 | Text only | General tasks |
| 4-Maverick | 8,192 tokens (~6,144 words) | 128,000 | Text only | General tasks |
Review Version Changes
Upgrade from 3.1-405B to 4-Maverick
- Longer responses — maximum response length increased from 2,048 tokens to 8,192 tokens (~6,144 words); useful for more detailed answers and longer content generation
Choose a Version
Latest version: 4-Maverick—Suitable for general-purpose tasks with moderate response length requirements.
