Models · Live catalog
Models
Every model.
Zero guesswork.
Nineteen free ways to get an answer, twelve frontier models on Pro or your own key. Every reply labels its exact model — this page says what each one is for.
Free pool · $0
Free models, named.
GPT-OSS 20B
Keyless model that answers when every provider is busy. No key, no account.
Open baseline · $0openai-fast
Free auto-router
Picks the best available $0 model per request. The default — leave it on.
Auto-router · $0openrouter/free
GLM 5.2
Strong all-round open chat model, good default for writing and Q&A.
Z.ai · chat · $0z-ai/glm-5.2:free
North Mini Code
Compact coding specialist for quick edits and reviews.
Cohere · code · $0cohere/north-mini-code:free
Qwen 3.8 27B
Capable mid-size model for chat, writing, and light code.
Alibaba · all-round · $0qwen/qwen3.8-27b:free
Nemotron 120B
Large reasoning model for hard multi-step problems.
Nvidia · reasoning · $0nvidia/nemotron-3-super-120b-a12b:free
Gemma 4 31B
Big instruction-tuned model for chat and analysis.
Google · all-round · $0google/gemma-4-31b-it:free
Gemma 4 26B
Smaller Gemma that answers fast for everyday tasks.
Google · efficient · $0google/gemma-4-26b-a4b-it:free
Nemotron Lightning
Speed-tuned model for quick drafts and iterations.
Nvidia · fast · $0nvidia/nemotron-3.5-lightning:free
Laguna Code S
Code-focused model for generation and debugging.
Poolside · code · $0poolside/laguna-s-2.1:free
Laguna Code XS
Smaller code model for instant completions and fixes.
Poolside · fast code · $0poolside/laguna-xs-2.1:free
Dots 3 Note
Note-style model for summaries and structured writing.
Dots Studio · notes · $0dots-studio/dots-3-note-preview:free
Ling Flash
Snappy model for chat and short-form answers.
InclusionAI · fast · $0inclusionai/ling-3.0-flash-fin:free
Ling Sante
Sibling fast model — the pool retries across both.
InclusionAI · fast · $0inclusionai/ling-3.0-flash-sante:free
LFM 2.5
Tiny model for instant, low-cost answers.
Liquid · tiny · $0liquid/lfm-2.5-2.6b:free
Nemotron Nano
Compact reasoner for step-by-step thinking.
Nvidia · reasoning · $0nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free
Nemotron Ultra
The biggest free model in the pool for heavy lifts.
Nvidia · largest · $0nvidia/nemotron-3-ultra-550b-a55b:free
Inkling
Conversational model for natural back-and-forth.
Thinking Machines · chat · $0thinkingmachines/inkling:free
Inkling Small
Smaller, quicker sibling for everyday chat.
Thinking Machines · fast · $0thinkingmachines/inkling-small:free
Pro shelf · $12/30 days or your key
Frontier models, named.
Claude Opus 5.5
Top-tier reasoning, coding, and long-context work.
Anthropic · flagship · Proanthropic/claude-opus-5.5
GPT-6 Luna Pro
Frontier generalist for hard builds and analysis.
OpenAI · flagship · Proopenai/gpt-6-luna-pro
Grok 4.7
Fast frontier model with strong reasoning.
xAI · flagship · Prox-ai/grok-4.7
Gemini 3.8 Flash
Frontier quality at flash speed and price.
Google · fast frontier · Progoogle/gemini-3.8-flash
DeepSeek V4 Pro
Pro-grade open reasoner for code and math.
DeepSeek · pro reasoning · Prodeepseek/deepseek-v4-pro
MiMo 2.6 Pro
MiMo flagship for chat, code, and agents.
Xiaomi · flagship · Proxiaomi/mimo-v2.6-pro
Qwen 3.8 Max
Largest Qwen tier for demanding tasks.
Alibaba · max tier · Proqwen/qwen3.8-max-prime
GPT-6 Sol Pro
Second GPT-6 pro flavor — compare outputs.
OpenAI · pro · Proopenai/gpt-6-sol-pro
Claude Opus 5
Previous Opus generation, still formidable.
Anthropic · pro · Proanthropic/claude-opus-5
Claude Fable 5.1
Storytelling-leaning Claude for writing flair.
Anthropic · creative · Proanthropic/claude-fable-5.1
Grok 4.6
Previous Grok generation at a lower price.
xAI · pro · Prox-ai/grok-4.6
Gemini 3.7 Flash
Previous flash generation, quick and cheap.
Google · fast · Progoogle/gemini-3.7-flash
Unlock with Pro or your own provider key — billed to you, never us. No credits behind a call? You get a free answer with an honest note.