DeepSeek V4.1 Flash 1017 entered the top 10
Ranked 6th, with 4.5% of traffic over the last 7 days.
What the data says. DeepSeek has 30.1% of tokens and 5.6% of estimated spend this week.
See DeepSeek V4.1 Flash 1017 on Model Season →DeepSeek V4.1 Flash 1017 entered the top 10
GLM 5.3 got 68% cheaper
From $1.26 to $0.41 per 1M tokens, blended. Input $0.079 → $0.079, output $4.80 → $1.40. New price since Oct 19.
What the data says. The model has 3.28% of traffic over the last 7 days, ranked 9th. Z.ai (GLM) has 12.1% of tokens and 2.7% of estimated spend this week. The lab's most used model is GLM 5.3 Flash, with 7.40% at $0.068 per 1M, and this one costs 6.0 times more.
See GLM 5.3 on Model Season →MiniMax M3.1 Flash (free) debuted in traffic
First volume recorded on Oct 19, 2026, 0 days after entering the catalog, with 0.65% of traffic over the last 7 days.
What the data says. The lab's most used model is MiniMax M3, with 0.87% at $0.53 per 1M.
See MiniMax M3.1 Flash (free) on Model Season →Mercury 3 is now in the catalog
Inception, $0.10 input and $0.35 output per 1M tokens, 1M context and optional reasoning. No traffic volume measured yet.
Hermes 3 405B Instruct is set to be deprecated on Nov 2, 2026
13 days left. It is the first notice in the catalog.
What the data says. No volume over the last 7 days. Its weekly peak was 1.19%.
See Hermes 3 405B Instruct on Model Season →Other changes
- DeprecationHermes 3 70B Instruct is set to be deprecated on Nov 2, 2026. 13 days left. It is the first notice in the catalog. See model →
- PriceGemma 4 26B A4B got 33% more expensive. From $0.11 to $0.14 per 1M tokens, blended. Input $0.068 → $0.090, output $0.23 → $0.30. New price since Oct 19. See model →
- PriceQwen3.5-27B got 12% cheaper. From $0.84 to $0.74 per 1M tokens, blended. Input $0.26 → $0.27, output $2.60 → $2.16. New price since Oct 19.
- PriceGLM 4.6 got 10% more expensive. From $0.88 to $0.96 per 1M tokens, blended. Input $0.50 → $0.55, output $2.00 → $2.20. New price since Oct 19. See model →
- ContextClef changed its context window. From 64k to 64k tokens.
Catalog snapshot of Oct 20, 2026, compared with the previous ones since Oct 13, 2026. Traffic through Oct 19, 2026. Price is a 75% input and 25% output blend, and only becomes news when the new value holds for two snapshots. Order follows a fixed relevance formula that weighs the type of change and the model’s share.