Model comparison
Two models, line by line — the winner is visible at a glance
Select two LLMs to generate a line-by-line comparison: input/output/cache-hit prices, time-based pricing, context windows, supported modalities and features, plus an example monthly cost using the exact formula of the cost calculator. For a whole-market comparison at identical usage and how to read it, see the pricing comparison guide.
What each comparison fixes
Every pair compares: input price, output price, cache-hit price, time-based pricing rules, context window and supported modalities, plus an example monthly cost identical to the cost calculator (100 calls/day · 2,000 input / 800 output tokens per call · off-peak rates · 30% cache hit rate). Prices are official published list prices (CNY or USD per 1M tokens; USD models are converted to CNY at the reference rate for comparison — indicative only); models without a published cache price use the miss price. The example monthly cost is an estimate at that fixed usage only — it is not a ranking of capability, output quality or overall value, and does not constitute purchase advice. Method and common mistakes: How to calculate LLM API costs.
Frequently compared pairs
· Budget lightweight tier: DeepSeek V4.1 Flash vs GLM-5.3-Flash DeepSeek V4.1 Flash vs Qwen3.8-Flash
· Flagship reasoning: DeepSeek V4 Pro vs GLM-5.3 GLM-5.3 vs Kimi K3
· Long context: DeepSeek V4.1 Flash vs Kimi K3 Kimi K3 vs Qwen3.8-Flash
· Same family, different tier: Qwen3.5-Plus vs Qwen3.7-Plus
· OpenAI, head to head: GPT-6 Luna vs GPT-6.1 Sol
· Anthropic, head to head: Claude Haiku 5.5 vs Claude Sonnet 5.5
· Cross-vendor lightweight: GLM-5.3-Flash vs GPT-6 Luna Claude Haiku 5.5 vs GLM-5.3-Flash Gemini 3.1 Flash-Lite vs Qwen3.8-Flash
· Cross-vendor workhorse: DeepSeek V4 Pro vs GPT-6.1 Sol Claude Sonnet 5.5 vs DeepSeek V4 Pro
All 136 pairs (every combination of 17 models) are reachable from the selector above; each comparison page has a fixed-basis summary and links to both model profiles.