Skip to content
Model Season every model has a seasonRadar
Radar · Oct 21, 2026

US orders federal civilian agencies to remove DeepSeek from their systems

28 itemsRSS

News × data

The story published out there and, below it, what real token traffic from the last 7 days says about who is in it. The headline opens the original article; the model name opens its page here.

Regulation ·

US orders federal civilian agencies to remove DeepSeek from their systems

The U.S. has ordered federal civilian agencies to stop buying and to remove DeepSeek products and third-party services built on its technology over national security concerns, according to the Department of Homeland Security. The order, also reported by Reuters, lets agencies request exceptions.

Launch ·

Google brings Gemini Personal Intelligence to Android Auto and cars

Google is bringing Personal Intelligence to Gemini in Android Auto and cars with Google built-in, according to the company. With the setting enabled, the assistant draws on Gmail, Calendar, Maps, Photos, YouTube and Search. It reaches Android Auto first, with cars with Google built-in coming in the next months.

Changes in the data

Price · DeepSeek

DeepSeek V4 Flash 0731 got 284% more expensive

From $0.17 to $0.66 per 1M tokens, blended. Input $0.016 → $0.440, output $0.64 → $1.32. New price since Oct 20.

What the data says. The model has 2.88% of traffic over the last 7 days, ranked 11th. DeepSeek has 30.1% of tokens and 6.7% of estimated spend this week. The lab's most used model is DeepSeek V4.1 Flash, with 16.69% at $0.13 per 1M, and this one costs 5.1 times more.

See DeepSeek V4 Flash 0731 on Model Season →
Price · DeepSeek

DeepSeek V4 Pro 0423 got 115% more expensive

From $0.26 to $0.56 per 1M tokens, blended. Input $0.21 → $0.45, output $0.42 → $0.90. New price since Oct 20.

What the data says. The model has 0.39% of traffic over the last 7 days, ranked 38th. DeepSeek has 30.1% of tokens and 6.7% of estimated spend this week. The lab's most used model is DeepSeek V4.1 Flash, with 16.69% at $0.13 per 1M, and this one costs 4.3 times more.

See DeepSeek V4 Pro 0423 on Model Season →
Catalog · Mistral AI

Voxtral Mini Transcribe 3 is now in the catalog

Mistral AI, $58.33 input and $0.0000 output per 1M tokens and 64k context. No traffic volume measured yet.

What the data says. The lab's most used model is Mistral Nemo, with 0.03% at $0.029 per 1M, and this one costs 1493 times more.

Catalog · Mistral AI

Voxtral TTS 3 is now in the catalog

Mistral AI, $13.00 input and $0.0000 output per 1M tokens and 4k context. No traffic volume measured yet.

What the data says. The lab's most used model is Mistral Nemo, with 0.03% at $0.029 per 1M, and this one costs 333 times more.

Other changes

  • CatalogVoxtral Small Transcribe 3 is now in the catalog. Mistral AI, $66.67 input and $0.0000 output per 1M tokens and 64k context. No traffic volume measured yet.
  • PriceClef got 471% more expensive. From $0.032 to $0.180 per 1M tokens, blended. Input $0.042 → $0.240, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven Flash v2 got 100% more expensive. From $15.00 to $30.00 per 1M tokens, blended. Input $20.00 → $40.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven Flash v2.5 got 100% more expensive. From $15.00 to $30.00 per 1M tokens, blended. Input $20.00 → $40.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven Multilingual v2 got 100% more expensive. From $30.00 to $60.00 per 1M tokens, blended. Input $40.00 → $80.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven Turbo v2 got 100% more expensive. From $15.00 to $30.00 per 1M tokens, blended. Input $20.00 → $40.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven Turbo v2.5 got 100% more expensive. From $15.00 to $30.00 per 1M tokens, blended. Input $20.00 → $40.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven v3 got 100% more expensive. From $30.00 to $60.00 per 1M tokens, blended. Input $40.00 → $80.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven v3 Conversational got 100% more expensive. From $15.00 to $30.00 per 1M tokens, blended. Input $20.00 → $40.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven v4 got 100% more expensive. From $30.00 to $60.00 per 1M tokens, blended. Input $40.00 → $80.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceEleven v4 Turbo got 100% more expensive. From $15.00 to $30.00 per 1M tokens, blended. Input $20.00 → $40.00, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceScribe v2 got 100% more expensive. From $22.92 to $45.83 per 1M tokens, blended. Input $30.56 → $61.11, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceScribe v2 Medical got 100% more expensive. From $22.92 to $45.83 per 1M tokens, blended. Input $30.56 → $61.11, output $0.0000 → $0.0000. New price since Oct 20.
  • PriceBeam 1.0 501B A23B got 68% cheaper. From $1.50 to $0.47 per 1M tokens, blended. Input $1.00 → $0.30, output $3.00 → $1.00. New price since Oct 20. See model →
  • PriceSolar Pro 4 got 67% more expensive. From $0.16 to $0.26 per 1M tokens, blended. Input $0.090 → $0.150, output $0.36 → $0.60. New price since Oct 20. See model →
  • PriceQwen3.8 27B got 50% cheaper. From $0.73 to $0.37 per 1M tokens, blended. Input $0.240 → $0.040, output $2.20 → $1.35. New price since Oct 20. See model →
  • PriceNemotron 3.5 Super VL got 36% more expensive. From $0.17 to $0.24 per 1M tokens, blended. Input $0.10 → $0.15, output $0.40 → $0.50. New price since Oct 20. See model →
  • PriceGemma 3 4B got 32% cheaper. From $0.063 to $0.043 per 1M tokens, blended. Input $0.050 → $0.040, output $0.100 → $0.050. New price since Oct 20. See model →
  • PriceQwen3.5 397B A17B got 31% cheaper. From $1.27 to $0.88 per 1M tokens, blended. Input $0.50 → $0.39, output $3.60 → $2.34. New price since Oct 20. See model →
  • PriceQwen3.6 35B A3B got 14% cheaper. From $0.36 to $0.31 per 1M tokens, blended. Input $0.15 → $0.10, output $1.00 → $0.95. New price since Oct 20. See model →

In the news: summaries written by AI using only what the cited sources published, reviewed before going live. The numbers in the “The data” strip come from our data, traffic from the last 7 days, never from the summary. Catalog snapshot of Oct 21, 2026, compared with the previous ones since Oct 14, 2026. Traffic through Oct 20, 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.