Models
Tell Zinley the job.
Auto picks the model.
The model can change from one job to the next. Your conversation and context stay together.
Claude
GPT
GrokDefault mode
Start with Auto.
Tell Zinley what needs to happen. Auto chooses the model for that job.
Everyday jobs stay efficient
Hard jobs get more depth
Your conversation stays together
32 active models
Choose manually when you want to.
Plus, Pro, and Max can choose models available on their plan. Most people can leave Auto on.
Claude
Anthropic
6 models
Claude
Anthropic
Models from Anthropic for reasoning, tool use, and long-running jobs.
Claude Opus 5.5
Coding, agent work, and computer use
Claude Fable 5.2
Long, reasoning-heavy jobs
Claude Opus 5.5 Fast
Faster Opus replies at a higher rate
Claude Fable 5.2 Fast
Faster Fable replies at a higher rate
Claude Sonnet 5.5
General work with tools
Claude Haiku 5.5
Quick answers and lighter jobs
GPT
OpenAI, via Microsoft Azure
4 models
GPT
OpenAI, via Microsoft Azure
Reasoning and multi-step coordination
GPT-5.6 Luna
OpenAI
Everyday work on every plan
GPT-5.6 Terra
OpenAI
Long-context reasoning
GPT-6 Astra
OpenAI
General high-capability work
GPT-6.1 Sol
OpenAI
General high-capability work
Muse
Meta
1 model
Muse
Meta
A Meta reasoning model for long, demanding jobs.
Muse Spark 1.3
Long, reasoning-heavy jobs
Grok
xAI
1 model
Grok
xAI
xAI models with live X (Twitter) search for current questions.
Grok 4.7
Real-time answers with live X (Twitter) search
Fugu
Sakana AI
1 model
Fugu
Sakana AI
A multi-agent system that coordinates one to three models and returns one answer.
Sakana Fugu Ultra
Multi-step work routed across specialist models
Open-weights
Fireworks AI
7 models
Open-weights
Fireworks AI
Open models served through US-hosted inference endpoints. Routing and retention depend on the provider configuration in use.
Kimi K3
Moonshot AI
Coding and agent work
Qwen 3.8 Max
Qwen (Alibaba)
General work with a large context window
GLM-5.3
Zhipu AI
Reasoning-heavy tool work
GLM-5.3 Flash
Zhipu AI
Quick answers and lighter jobs
MiniMax M3
MiniMax
Fast, economical agentic runs
MiniMax M3.1 Flash
MiniMax
Fast, economical agentic runs
DeepSeek V4.1 Flash
DeepSeek
Large codebases and documents
Images, video & speech
Google, OpenAI and xAI
12 models
Images, video & speech
Google, OpenAI and xAI
Models for images, video, and speech.
Nano Banana Pro
Google
Detailed image generation and editing
Nano Banana 2
Google
Default image generation and editing
Nano Banana Lite
Google
Quick, economical image work
GPT Image 2
OpenAI
Alternate image style; automatic fallback
GPT Image 1.5
OpenAI
Alternate image style; automatic fallback
Grok Imagine Quality
xAI
Detailed images and text rendering
Grok Imagine
xAI
Lower-cost image generation
Veo 3.1
Google
Video generation with audio
Veo 3.1 Fast
Google
Quick video drafts
Grok Imagine Video
xAI
Video with native audio, up to 15 seconds
Grok Imagine Video 1.5
xAI
Motion from a text prompt or image
Speech-to-Text
OpenAI
Speech-to-text for calls and voice notes
