Previous generation models with the largest context window
Rule: Sorted by context window from the maker's docs, largest first, then by max output. This ranks one published spec. It is not a quality test.
| # | Name | Context window | ||
|---|---|---|---|---|
| 1 |  Gemini 2.5 FlashGoogle | 1,048,576 input tokens | ||
| 2 | Gemini 2.5 Flash-LiteGoogle | 1,048,576 input tokens | ||
| 3 | Gemini 2.5 ProGoogle | 1,048,576 input tokens | ||
| 4 |  Claude Opus 4.6Anthropic | 1M tokens | ||
| 5 | Claude Opus 4.7Anthropic | 1M tokens | ||
| 6 | Claude Opus 4.8Anthropic | 1M tokens | ||
| 7 | Claude Opus 5Anthropic | 1M tokens | ||
| 8 | Claude Sonnet 4.6Anthropic | 1M tokens | ||
Get previous generation price changes by email | ||||
| 9 | Claude Sonnet 5Anthropic | 1M tokens | ||
| 10 | Kimi K2.6Moonshot AI | 262,144 tokens | ||
| 11 | Claude Haiku 4.5Anthropic | 200K tokens | ||
| 12 | Claude Opus 4.5Anthropic | 200K tokens | ||
| 13 | Claude Sonnet 4.5Anthropic | 200K tokens | ||
| 14 | GLM-4.7Z.ai | 200K tokens | ||
| 15 | GLM-5Z.ai | 200K tokens | ||
| 16 | GLM-4.5-AirZ.ai | 128K tokens | ||
Get an email when a previous generation AI we track changes its price
Questions
Which previous generation model has the largest context window
Gemini 2.5 Flash, 1,048,576 input tokens, per ai.google.dev.