fineprint what AI companies quietly changed · RSS

Groq

Every recorded change to 3 public pages. Faded entries are wording-only changes.

October 2026
  1. Oct 16Added Muse Spark Open model with a 350 input price and ContactSales output pricing.modelspricing+1 −0raw difflive page
  2. Oct 161 line(s) added, 0 removed.modelspricing+1 −0raw difflive page
  3. Oct 06Reduced the requests per minute for openai/gpt-oss-safeguard-20b from 30 to 3 and the tokens per minute from 8K to 2K.rate-limitslimits+1 −1raw difflive page
  4. Oct 061 line(s) added, 1 removed.rate-limitslimits+1 −1raw difflive page
September 2026
  1. Sep 29Added deprecation notices for groq/compound, groq/compound-mini, and qwen/qwen3.6-27b with shutdown dates and recommended replacements.deprecations+9 −0raw difflive page
  2. Sep 299 line(s) added, 0 removed.deprecations+9 −0raw difflive page
  3. Sep 24Removed Groq Compound and Qwen/Qwen3.6-27B models and updated the context window for Qwen/Qwen3.8-27B to 131,072 tokens.modelspricing+2 −16raw difflive page
  4. Sep 242 line(s) added, 16 removed (around "Groq Compound").modelspricing+2 −16raw difflive page
  5. Sep 23Removed groq/compound and groq/compound-mini models from the rate limits table.rate-limitslimits+0 −2raw difflive page
  6. Sep 230 line(s) added, 2 removed.rate-limitslimits+0 −2raw difflive page
  7. Sep 16Removed qwen/qwen3.6-27b from the rate limits table.rate-limitslimits+0 −1raw difflive page
  8. Sep 160 line(s) added, 1 removed.rate-limitslimits+0 −1raw difflive page
  9. Sep 05Added qwen/qwen3.8-27b to the rate limits table with a 30 RPM, 1K RPM, 8K TPM, and 200K TDP limit.rate-limitslimits+1 −0raw difflive page
  10. Sep 051 line(s) added, 0 removed.rate-limitslimits+1 −0raw difflive page
  11. Sep 05Added Llama 3.1 8B Enterprise, Llama 3.3 70B Enterprise, and Qwen/Qwen3.8-27B models to the pricing table.modelspricing+3 −0raw difflive page
  12. Sep 053 line(s) added, 0 removed.modelspricing+3 −0raw difflive page
August 2026
  1. Aug 25Removed llama-3.1-8b-instant and llama-3.3-70b-versatile from the rate limits table.rate-limitslimits+0 −2raw difflive page
  2. Aug 250 line(s) added, 2 removed.rate-limitslimits+0 −2raw difflive page
  3. Aug 22Removed Llama 3.1 8B and Llama 3.3 70B models from the pricing table.modelspricing+0 −2raw difflive page
  4. Aug 220 line(s) added, 2 removed.modelspricing+0 −2raw difflive page
July 2026
  1. Jul 24Removed Llama 4 Scout 17B 16E, Qwen3-32B, and Qwen/Qwen3.6-27B while adding MiniMax M2.7 and updating Qwen/Qwen3.6-27B context limits.modelspricing+2 −3raw difflive page
  2. Jul 242 line(s) added, 3 removed.modelspricing+2 −3raw difflive page
  3. Jul 23Removed allam-2-7b, meta-llama/llama-4-scout-17b-16e-instruct, and qwen/qwen3-32b from the rate limits table.rate-limitslimits+1 −4raw difflive page
  4. Jul 231 line(s) added, 4 removed.rate-limitslimits+1 −4raw difflive page
  5. Jul 03Announced the deprecation of llama-3.1-8b-instant, llama-3.3-70b-versatile, qwen/qwen3-32b, and meta-llama/llama-4-scout-17b-16e-instruct with specific shutdown dates and recommended replacements.deprecations+10 −0raw difflive page
  6. Jul 0310 line(s) added, 0 removed.deprecations+10 −0raw difflive page
June 2026
  1. Jun 23Added pricing, rate limits, and technical specifications for production and preview models including Llama 3.1, Llama 3.3, GPT OSS, and Whisper.modelspricing+22 −3raw difflive page
  2. Jun 2322 line(s) added, 3 removed.modelspricing+22 −3raw difflive page
  3. Jun 23Added qwen/qwen3.6-27b to the rate limits table with 30 requests per minute and 200K context length.rate-limitslimits+1 −0raw difflive page
  4. Jun 231 line(s) added, 0 removed.rate-limitslimits+1 −0raw difflive page
  5. Jun 15Removed Llama 3.1, Llama 3.3, GPT OSS, Whisper, Compound, and several preview models from the pricing page.modelspricing+3 −21raw difflive page
  6. Jun 153 line(s) added, 21 removed.modelspricing+3 −21raw difflive page
May 2026
  1. May 17Added ITPM and OTPM rate limits and a breakdown of input and output tokens for organizations with separate per-minute limits.rate-limitslimits+4 −0raw difflive page
  2. May 174 line(s) added, 0 removed.rate-limitslimits+4 −0raw difflive page
Pages watched