Catalog
Open a model. The chat is already on it.
22 models from 12 providers. Desk rates are what the pool debits per million tokens. Holders spend them inside a daily allowance. Vendor invoices can differ once live routing is on.
Holder$0.25 /day
hold $10 · standard
1 API key
Pro$1.50 /day
hold $100 · standard, premium, media
3 API keys
Whale$5.00 /day
hold $250 · standard, premium, media
10 API keys
| Model | Class | Context | Input / 1M | Output / 1M | |
|---|---|---|---|---|---|
| Claude Sonnetanthropic/claude-sonnetLong-context writing and careful code review. | premium | 1M | $3.00 | $15.00 | Open |
| Claude Opusanthropic/claude-opusThe slow, thorough pass when the draft has to hold. | premium | 1M | $15.00 | $75.00 | Open |
| Claude Haikuanthropic/claude-haikuShort turns, low debit, still precise. | standard | 200K | $0.80 | $4.00 | Open |
| GPTopenai/gptGeneral desk model for mixed questions. | premium | 400K | $2.50 | $10.00 | Open |
| GPT miniopenai/gpt-miniEveryday prompts inside a Holder allowance. | standard | 400K | $0.40 | $1.60 | Open |
| GPT Imageopenai/gpt-imageLogo and mark drafts for a launch. | media | 128K | $5.00 | $15.00 | Open |
| Gemini Progoogle/gemini-proWide context for docs, threads, and filings. | premium | 1M | $2.00 | $12.00 | Open |
| Gemini Flashgoogle/gemini-flashFast reads over long pasted text. | standard | 1M | $0.30 | $1.20 | Open |
| Gemini Imagegoogle/gemini-imageStill images from a short brief. | media | 128K | $2.00 | $8.00 | Open |
| Grokx-ai/grokDirect answers with less ceremony. | premium | 500K | $2.00 | $6.00 | Open |
| Grok Fastx-ai/grok-fastA lighter Grok for the Holder tier. | standard | 256K | $0.40 | $1.20 | Open |
| DeepSeek Chatdeepseek/deepseek-chatCheap reasoning for drafts and diffs. | standard | 1M | $0.27 | $1.10 | Open |
| DeepSeek Reasonerdeepseek/deepseek-reasonerExtra steps before it commits to an answer. | premium | 128K | $0.55 | $2.20 | Open |
| Qwen Plusqwen/qwen-plusBilingual drafting, including Chinese briefs. | standard | 1M | $0.40 | $1.20 | Open |
| Qwen Maxqwen/qwen-maxHeavier Qwen when the brief is messy. | premium | 256K | $1.60 | $6.40 | Open |
| Mistral Largemistral/mistral-largeEuropean desk model for tight prose. | premium | 128K | $2.00 | $6.00 | Open |
| Mistral Smallmistral/mistral-smallSmall debit, good for rewrites. | standard | 128K | $0.20 | $0.60 | Open |
| Llamameta/llamaOpen-weight style answers at a low rate. | standard | 128K | $0.18 | $0.60 | Open |
| Commandcohere/commandRetrieval-shaped answers and summaries. | standard | 128K | $0.50 | $1.50 | Open |
| Sonarperplexity/sonarDesk notes that expect a source list. | standard | 128K | $1.00 | $1.00 | Open |
| Kimimoonshot/kimiLong Chinese and English documents. | standard | 256K | $0.60 | $2.50 | Open |
| GLMz-ai/glmA second opinion on a launch writeup. | standard | 200K | $0.40 | $1.40 | Open |