Kolbo.AIKolbo.AI Docs
Models

Chat Models

Every chat and language model available on Kolbo.AI, plus Smart Select auto-routing

The Kolbo.AI Chat tool gives you the leading language models from every major lab in one conversation, with memory, custom agents, artifacts, and web search. Switch models mid-thread, or let Smart Select pick the best one for you.

Available Models

ModelProviderContextReasoning effortBest for
GPT (latest)OpenAIUp to 1M tokensConfigurableLong-context reasoning and agentic workflows
Claude (Opus and Sonnet)AnthropicUp to 1M tokensHigh-quality reasoningComplex reasoning, agentic coding, multi-step tasks
Gemini (Flash and Pro)GoogleUp to 1M tokensMultimodalLong-context multimodal analysis and coding
Grok 4.xxAI1M to 2M tokensConfigurableLow-hallucination agentic tool calling and strict instruction following
DeepSeekDeepSeekLarge contextReasoning variantsCoding, multi-step reasoning, cost-effective frontier
QwenAlibabaUp to 1M tokensConfigurableAgent-centric coding and long-horizon tasks
Smart SelectKolbo (auto-routing)Inherits chosen modeln/aAutomatically picks the best model for each turn

Web search is a platform-level feature, available on any chat model rather than a separate model. When enabled, the chat synthesizes web-grounded answers with numbered inline citations and a source list, and supports search in multiple languages. See Chat Tools for details.

Reasoning Effort

Several models (such as Grok 4.x and the latest GPT and Qwen) expose a reasoning effort control. Lower effort is faster and cheaper for simple tasks, higher effort spends more time thinking for complex problems.

Smart Select

Smart Select removes the guesswork. Instead of choosing a model yourself, Smart Select reads your prompt and routes it to the model best suited to the task, balancing quality, speed, and cost automatically.

Where to use these

These models power the Chat tool and the Smart Agent. See Chat Tools for memory, custom agents, artifacts, folders, and PDF export.