LLM model catalog
30 models from 9 vendors, with context windows, output limits and published list prices, last verified Oct 15, 2026.
| Model | Vendor | Input / 1M | Output / 1M | Context | Max output |
|---|---|---|---|---|---|
| Qwen3.7-Plus | Alibaba Cloud (Qwen) | $0.300 | $1.19 | 1M | — |
| Qwen3.8-Flash | Alibaba Cloud (Qwen) | $0.120 | $0.400 | 1M | — |
| Qwen3.8-Max | Alibaba Cloud (Qwen) | $1.79 | $5.37 | 1M | — |
| Claude Fable 5.1 | Anthropic | $10 | $50 | 1M | 128K |
| Claude Haiku 5.5 | Anthropic | $0.100 | $0.500 | 1M | 128K |
| Claude Opus 5.5 | Anthropic | $4 | $20 | 1M | 128K |
| Claude Sonnet 5.5 | Anthropic | $2 | $10 | 1M | 128K |
| DeepSeek-V4-Pro-0813 | DeepSeek | $1.32 | $3.96 | 1M | 384K |
| DeepSeek-V4.1-Flash | DeepSeek | $0.300 | $1.20 | 1M | 384K |
| Gemini 3.1 Pro Preview | $2 | $12 | 1.0M | 65.5K | |
| Gemini 3.8 Flash | $0.750 | $3.75 | 1.0M | 65.5K | |
| Gemini 3.8 Flash Cyber | $1.50 | $7.50 | 1.0M | 65.5K | |
| Gemini 4 Argon | $2 | $10 | 1.0M | 1.0M | |
| Muse Wonder | Meta | $4 | $20 | 1M | — |
| Kimi K2.6 | Moonshot AI | $0.950 | $4 | 262.1K | — |
| Kimi K2.7 Code | Moonshot AI | $0.950 | $4 | 262.1K | — |
| Kimi K3 | Moonshot AI | $3 | $15 | 1.0M | — |
| GPT-5.3-Codex | OpenAI | $1.75 | $14 | 400K | 128K |
| GPT-5.6 Sol | OpenAI | $4 | $20 | 1.1M | 128K |
| GPT-6 Astra | OpenAI | $10 | $50 | 1.1M | 128K |
| GPT-6 Luna | OpenAI | $0.100 | $0.500 | 1.1M | 128K |
| GPT-6.1 Sol | OpenAI | $2 | $10 | 1.1M | 128K |
| Grok 4.3 | xAI | $1.25 | $2.50 | 1M | — |
| Grok 4.5 | xAI | $2 | $6 | 500K | — |
| Grok 4.6 | xAI | $2 | $6 | 500K | — |
| Grok 4.7 | xAI | $2 | $6 | 500K | — |
| Grok Build 0.1 | xAI | $1 | $2 | 256K | — |
| GLM-5.3 | Zhipu AI | $1.19 | $4.18 | 1M | — |
| GLM-5.3-Flash | Zhipu AI | $0.120 | $0.420 | 1M | — |
| GLM-5.3-FlashX | Zhipu AI | $0.300 | $1.04 | 1M | — |
Prices are USD per million tokens, from each vendor's published pricing page. Click a model for its source link and worked cost examples.
Choosing on price alone
The cheapest output price in the catalog is Qwen3.8-Flash at $0.400 per million output tokens; the most expensive is GPT-6 Astra at $50. That spread matters far more than the input side for chat-shaped workloads, where output tokens are a small share of tokens and a large share of the bill.
Price is only one axis. Small models fail differently from large ones, and a cheap model that needs two attempts is not cheap. Use the cost calculator for the money, then your own evaluation for the quality.
By vendor
Keep this table in your own systems
Pro gives you the catalog as JSON and CSV, plus a bulk cost-estimation endpoint.
Pro pricing API docs