Models
The CLI lets you choose a model, but the server remains the authority on whether a model ID is supported. The default chat model is google/gemini-3.6-flash.
Model
Provider
Cost
Thinking
deepseek/deepseek-v4-flash-0731
deepseek
$
Yes
qwen/qwen3.7-flash
qwen
$
Yes
stepfun/step-3.5-flash
stepfun
$
Yes
openai/gpt-5-nano
openai
$
Yes
qwen/qwen3-coder-flash
qwen
$
No
poolside/laguna-s-2.1
poolside
$
Yes
inclusionai/ling-3.0-flash
inclusionai
$
No
qwen/qwen3.5-flash-02-23
qwen
$
No
bytedance-seed/seed-1.6-flash
bytedance-seed
$
Yes
qwen/qwen3-coder-30b-a3b-instruct
qwen
$
No
google/gemini-3.5-flash-lite
google
$$
Yes
qwen/qwen3.8-max
qwen
$$
Yes
qwen/qwen3.7-max
qwen
$$
Yes
google/gemini-3.6-flash
google
$$
Yes
qwen/qwen3.5-122b-a10b
qwen
$$
No
qwen/qwen3.6-27b
qwen
$$
No
openai/gpt-5-mini
openai
$$
Yes
openai/gpt-5.4-mini
openai
$$
Yes
z-ai/glm-5
z-ai
$$
Yes
moonshotai/kimi-k2.5
moonshotai
$$
Yes
google/gemini-3.5-flash
google
$$$
Yes
openai/gpt-5
openai
$$$
Yes
openai/gpt-5.2
openai
$$$
Yes
openai/gpt-5.3
openai
$$$
Yes
openai/gpt-5.4
openai
$$$
Yes
anthropic/claude-sonnet-4.5
anthropic
$$$
Yes
anthropic/claude-sonnet-4.6
anthropic
$$$
Yes
google/gemini-3.1-pro-preview
google
$$$
Yes
moonshotai/kimi-k3
moonshotai
$$$
Yes
minimax/minimax-01
minimax
$$$
No
anthropic/claude-opus-4.6
anthropic
$$$$
Yes
anthropic/claude-opus-4.7
anthropic
$$$$
Yes
anthropic/claude-opus-4.8
anthropic
$$$$
Yes
anthropic/claude-opus-5
anthropic
$$$$
Yes
anthropic/claude-opus-5-fast
anthropic
$$$$
Yes
anthropic/claude-opus-4.8-fast
anthropic
$$$$
Yes
anthropic/claude-opus-4.7-fast
anthropic
$$$$
Yes
openai/gpt-5.5
openai
$$$$
Yes
openai/gpt-5.5-pro
openai
$$$$
Yes
openai/gpt-5.4-pro
openai
$$$$
Yes
What do the cost tiers mean?
Cost tiers are based on the model's output-token price per million tokens. The boundaries are fixed, so the tiers never overlap.
Input and output prices are still shown on each model's pricing data. The tier itself is determined by output cost because agentic coding workloads can generate substantial output.
$0–<800 credits output (per million tokens)
$$800–<3,200 credits output (per million tokens)
$$$3,200–<10,000 credits output (per million tokens)
$$$$10,000+ credits output (per million tokens)
Credits are what you are actually charged. See Dashboard & billing for how credits convert to Naira for each credit pack.
Model picker search
The model picker supports searching by model ID, provider name, cost tier ($, $$, $$$, $$$$) and thinking-related keywords like thinking or think.