dotpoin-composer
Auto
Picks the right model for the job: chat, tools, and vision
$1 = 1,000,000 tokens promo test price
Pricing & models
Pay-as-you-go: chat — input, output & cache (USD / 1M tokens) plus the “$1 = N” guide; images (flat per request) and monthly n8n. One DotPoin balance.
DotPoin
Chat and image generation in one place.
DotPoin
LLM
Image
dotpoin-image
Image
Image generation and editing
$0.05 per request
dotpoin-image-uncensored
Uncensored
Uncensored image generation and editing
$0.10 per request
MiMo v2.5 — chat, agents & speech
Chat with vision, agents, speech synthesis and recognition — one key.
mimo-v2.5
Chat
Chat & full vision (images)
$1 = 3,333,333 tokens incoming cache
$1 = 1,666,666 tokens incoming, non-cache
$1 = 1,000,000 tokens outgoing
mimo-v2.5-pro
Agents
Agents, coding, deep reasoning
$1 = 2,000,000 tokens incoming cache
$1 = 1,000,000 tokens incoming, non-cache
$1 = 400,000 tokens outgoing
mimo-v2.5-tts
TTS
Speech synthesis
$1 = 1,000,000 tokens
mimo-v2.5-tts-voicedesign
Design
Voice from text description
$1 = 1,000,000 tokens
mimo-v2.5-tts-voiceclone
Clone
Voice clone from sample
$1 = 1,000,000 tokens
mimo-v2.5-asr
ASR
Speech recognition
$1 = 1,000,000 tokens
DeepSeek V4
Chat, coding, and hard tasks — Flash with vision, Pro with Reasoner.
deepseek-flash
Chat, tools & vision (V4.1)
$1 = 6,666,666 tokens incoming cache
$1 = 1,428,571 tokens incoming, non-cache
$1 = 333,333 tokens outgoing
deepseek-v4-flash
Fast chat & tool use (legacy — support ending soon, use deepseek-flash)
$1 = 2,000,000 tokens incoming cache
$1 = 1,000,000 tokens incoming, non-cache
$1 = 333,333 tokens outgoing
deepseek-v4-pro
Deep reasoning (Reasoner)
$1 = 1,250,000 tokens incoming cache
$1 = 333,333 tokens incoming, non-cache
$1 = 166,666 tokens outgoing
DeepSeek V3 · Compatibility
Classic deepseek-chat / coder / reasoner IDs — no client code changes.
Runtime: Xiaomi MiMo v2.5 on DotPoin. LiteLLM billing matches mimo-v2.5 ($1 = 1,000,000).
deepseek-chat
Chat
Classic deepseek-chat · MiMo v2.5 without reasoning · chat, vision & tools
$1 = 3,333,333 tokens incoming cache
$1 = 1,666,666 tokens incoming, non-cache
$1 = 1,000,000 tokens outgoing
deepseek-coder
Coder
Classic deepseek-coder · MiMo v2.5 without reasoning · code & tool-use
$1 = 3,333,333 tokens incoming cache
$1 = 1,666,666 tokens incoming, non-cache
$1 = 1,000,000 tokens outgoing
deepseek-reasoner
Reasoner
Classic deepseek-reasoner · MiMo v2.5 with reasoning · complex tasks
$1 = 3,333,333 tokens incoming cache
$1 = 1,666,666 tokens incoming, non-cache
$1 = 1,000,000 tokens outgoing
Other models
Catalog: cheap everyday and strong models (Kimi, GLM, Qwen, MiniMax, gpt-oss).
Kimi-K2.5
Kimi
Long context, vision, tools, reasoning
$1 = 1,428,571 tokens incoming cache
$1 = 1,111,111 tokens incoming, non-cache
$1 = 222,222 tokens outgoing
Kimi-K3
Kimi · flag
Kimi flagship: 1M context, vision, tools, reasoning
$1 = 370,370 tokens incoming cache
$1 = 185,185 tokens incoming, non-cache
$1 = 37,037 tokens outgoing
GLM-5.3-Flash
Fast
Fast chat, vision, tools
$1 = 3,333,333 tokens incoming cache
$1 = 3,333,333 tokens incoming, non-cache
$1 = 1,000,000 tokens outgoing
GLM-5.3
Hard
Strong general / agents
$1 = 384,615 tokens incoming cache
$1 = 357,142 tokens incoming, non-cache
$1 = 113,636 tokens outgoing
Qwen3.8-2.4T-A95B
Flagship
Flagship MoE: coding & agents
$1 = 400,000 tokens incoming cache
$1 = 250,000 tokens incoming, non-cache
$1 = 83,333 tokens outgoing
MiniMax-M3
Mid
Multimodal mid-tier
$1 = 1,666,666 tokens incoming cache
$1 = 1,666,666 tokens incoming, non-cache
$1 = 416,666 tokens outgoing
gpt-oss-20b
Cheap
Cheap everyday / drafts
$1 = 12,500,000 tokens incoming, non-cache
$1 = 2,500,000 tokens incoming cache
$1 = 2,777,777 tokens outgoing
gpt-oss-120b
Cheap+
Cheap, a bit stronger than 20b
$1 = 10,000,000 tokens incoming, non-cache
$1 = 2,000,000 tokens incoming cache
$1 = 1,111,111 tokens outgoing
Qwen3.5-9B
Light
Light thinking + tools + vision
$1 = 5,000,000 tokens incoming, non-cache
$1 = 1,000,000 tokens incoming cache
$1 = 3,333,333 tokens outgoing
Personal n8n
Your isolated automation server on a dedicated VPS — connect services, APIs, and AI workflows without writing backend code.
Servers in Europe and the Americas — many LLMs and APIs are blocked in some countries; this location usually avoids those geo-limits.
DotPoin models (chat and images), MiMo, DeepSeek, and other models connect out of the box. Creating a server charges from your DotPoin balance. Or connect your own provider accounts inside n8n.