AI - Model Reference
Canonical model list for AI nodes, including deprecated model fallback mappings.
Chat Message Models
Used by AI - Chat Message.
OpenAI
Model | <abbr title="Image input support">🖼️</abbr> | <abbr title="Video input support">🎥</abbr> | <abbr title="Audio input support">🔊</abbr> | Best Suited For |
|---|---|---|---|---|
| ✅ | - | - | Most capable OpenAI model for the hardest end-to-end work. |
| ✅ | - | - | Flagship GPT-6 tier for complex professional work, coding, and multi-step reasoning. |
| ✅ | - | - | Cheapest, fastest GPT-6 tier for cost-sensitive, high-volume workloads. |
| ✅ | - | - | Flagship GPT-5.6 tier for complex professional work, coding, and multi-step reasoning. |
| ✅ | - | - | Balanced GPT-5.6 tier trading some quality for lower cost on everyday work. |
| ✅ | - | - | Cheapest, fastest GPT-5.6 tier for cost-sensitive, high-volume workloads. |
| ✅ | - | - | Highest-quality OpenAI option for complex coding and multi-step reasoning. |
| ✅ | - | - | Balanced flagship GPT-5.4 option when you need stronger quality than mini/nano tiers. |
| ✅ | - | - | Fast, efficient GPT-5.4 tier for coding, subagents, and high-volume workloads. |
| ✅ | - | - | Smallest, cheapest GPT-5.4 for classification, extraction, and coding subagents. |
| ✅ | - | - | High-quality general reasoning and complex multi-step tasks. |
| ✅ | - | - | Stable pinned GPT-5 behaviour for production workflows. |
| ✅ | - | - | Lower-latency GPT-5 tier for routine assistant tasks. |
| ✅ | - | - | Very fast, lightweight completions and high-throughput automation. |
| ✅ | - | - | Reliable balanced model for broad enterprise use cases. |
| ✅ | - | - | Cost-sensitive general chat and classification workloads. |
| ✅ | - | - | Ultra-fast short responses and basic transformations. |
| ✅ | - | - | Efficient reasoning tasks where speed and cost both matter. |
| ✅ | - | - | Strong multimodal text and image understanding with broad compatibility. |
| ✅ | - | - | Budget-friendly multimodal interactions and assistant features. |
| - | - | ✅ | General audio input/output through Chat Completions. |
| - | - | ✅ | Cost-efficient audio input/output through Chat Completions. |
Anthropic
Model | <abbr title="Image input support">🖼️</abbr> | <abbr title="Video input support">🎥</abbr> | <abbr title="Audio input support">🔊</abbr> | Best Suited For |
|---|---|---|---|---|
| ✅ | - | - | Highest-capability Claude tier for demanding reasoning and long-horizon agentic work. |
| ✅ | - | - | Highest-capability Claude tier for long-running agents and the hardest reasoning work. |
| ✅ | - | - | Current flagship Opus for agentic coding and enterprise work. |
| ✅ | - | - | Prior flagship Opus for complex agentic coding and enterprise work. |
| ✅ | - | - | Prior flagship Opus for highest-quality autonomous coding and long-horizon reasoning workflows. |
| ✅ | - | - | Current flagship Sonnet tier for production copilots, fast agentic workflows, and balanced cost. |
| ✅ | - | - | Prior flagship Opus with strong deep-reasoning quality and broad agentic coding capability. |
| ✅ | - | - | Earliest 4.x Opus baseline in this list, useful for compatibility-sensitive production workloads. |
| ✅ | - | - | Best combination of speed and intelligence for production copilots. |
| ✅ | - | - | Deep analysis, nuanced writing, and difficult reasoning tasks. |
| ✅ | - | - | Balanced quality and speed for most production copilots. |
| ✅ | - | - | Fast lightweight Claude option for shorter responses. |
Google AI
Model | <abbr title="Image input support">🖼️</abbr> | <abbr title="Video input support">🎥</abbr> | <abbr title="Audio input support">🔊</abbr> | Best Suited For |
|---|---|---|---|---|
| ✅ | ✅ | ✅ | Google's flagship reasoning model for complex coding, science, and autonomous multi-step tasks. |
| ✅ | ✅ | ✅ | Most intelligent Gemini Flash model for long-horizon software engineering and autonomous agents. |
| ✅ | ✅ | ✅ | High-speed Flash model for everyday coding, agentic tool use, and multi-step execution. |
| ✅ | ✅ | ✅ | Previous-generation Flash model balancing speed and multimodal capability. |
| ✅ | ✅ | ✅ | Cost-efficient model for high-volume agentic tasks, translation, and simple data processing. |
| ✅ | ✅ | ✅ | Stable Gemini 3.x chat model for agentic workflows, coding, and long-context assistants. |
| ✅ | ✅ | ✅ | Lowest-cost current Gemini 3.x option for high-volume classification, extraction, routing, and lightweight multimodal tasks. |
| ✅ | ✅ | ✅ | Advanced multimodal reasoning with strong media handling. |
| - | ✅ | ✅ | Cost-efficient model for lightweight multimodal workloads. |
Groq
Model | <abbr title="Image input support">🖼️</abbr> | <abbr title="Video input support">🎥</abbr> | <abbr title="Audio input support">🔊</abbr> | Best Suited For |
|---|---|---|---|---|
| - | - | - | Open model inference with strong text quality on Groq. |
| - | - | - | Faster open model option for practical general chat. |
| - | - | - | Safety-sensitive open model scenarios needing stricter behaviour. |
| - | - | - | High-quality Groq-hosted general-purpose text generation. |
Capability notes:
✅indicates the model supports that input type.-indicates it does not.- Google:
gemini-4-argonuses medium thinking by default. Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and later Gemini models do not support Temperature, Top K, or Top P. - Google:
gemini-4-argonis USD $4/$20 per 1M input/output tokens. - Anthropic:
claude-fable-5-1,claude-fable-5,claude-opus-5-5,claude-opus-5,claude-opus-4-8,claude-opus-4-7, andclaude-sonnet-5use adaptive thinking and do not support Temperature, Top K, Top P, or stop sequences. - OpenAI flagship pricing reference:
gpt-5.5is 2xgpt-5.4on standard token rates. - OpenAI pricing reference:
gpt-6-astrais $10 / $50 per 1M input/output tokens, with cached input at $1 / 1M tokens. - OpenAI pricing reference:
gpt-6-solis $2 / $10 per 1M input/output tokens, with cached input at $0.20 / 1M tokens. - OpenAI pricing reference:
gpt-6-lunais $0.10 / $0.50 per 1M input/output tokens, with cached input at $0.01 / 1M tokens. - OpenAI pricing reference:
gpt-5.6-solis $4 / $20 per 1M input/output tokens; cached input is $0.40 / 1M tokens. - Google deprecation:
gemini-3-flash-previewis deprecated in favour ofgemini-3.5-flash. - Google deprecation:
gemini-2.5-proandgemini-2.5-flashare deprecated. Starting December 1, 2026, Google routes Pro requests togemini-3.8-flashand Flash requests togemini-3.5-flash-lite.gemini-2.5-flash-literemains available with no announced shutdown date.
Data Extraction Models
Used by AI - Data Extraction.
- Uses the same models as AI - Chat Message.
- For exact model names, use the Chat Message list above.
Compute Embedding Models
Used by AI - Compute Embedding, AI - Vector Context Add, and AI - Vector Context Query.
text-embedding-ada-002(OpenAI, 1536 dimensions)cohere.embed-multilingual-v3(Cohere, 1024 dimensions)cohere.embed-english-v3(Cohere, 1024 dimensions)
Image Generate Models
Used by AI - Image Generate.
openai:gpt-image-2(OpenAI GPT Image 2, the default)
Text to Speech Models
Used by AI - Text to Speech.
openai:tts-1openai:tts-1-hd
Transcribe Models
Used by AI - Transcribe.
openai:whisper-1
Deprecated Models and Fallbacks
Chat Message Deprecated Models
These model names are accepted for import and backward compatibility. Each has a replacement listed below.
Deprecated | Replacement |
|---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Image Generate Deprecated Models
Deprecated | Replacement |
|---|---|
|
|
|
|
|
|