References 683% CONFIDENCEOverall confidence: 83%How well the pin's source and references back up its dates.Weighted average of how firmly 6 references, the source included, support the pin's start and end times; a reference counts half as much for every 180 days older than the newestShow all pins at 75% confidence or better

The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.

• [1]90%
  deepseek.com/en/news/r1-0528

  deepseek.com· Posted Oct 12, 2026· Starts May 28, 2025 ✓· 19% of score

  DeepSeek's^[3] post "DeepSeek-R1-0528 Release" is dated May 28, 2025 and says "DeepSeek-R1-0528 is here!"; its API changelog lists "Date: 2025-05-28 deepseek-reasoner upgraded to DeepSeek-R1-0528" and the Hugging Face^[2] repo was created the same day.

• [2]90%
  deepseek-ai/DeepSeek-R1-0528 - Hugging Face

  huggingface.co· Added Oct 12, 2026· 19% of score

  DeepSeek's^[1][3] model card: a "minor version upgrade" with more compute and post-training algorithm changes, approaching O3 and Gemini 2.5 Pro, the full R1-versus-R1-0528 table, system-prompt support and the distilled DeepSeek-R1-0528-Qwen3-8B.

• [3]90%
  Change Log - DeepSeek API Docs

  api-docs.deepseek.com· Added Oct 12, 2026· 19% of score

  DeepSeek's^[1] changelog entry for 2025-05-28: deepseek-reasoner upgraded to R1-0528 with AIME 2025 70.0 to 87.5, GPQA 71.5 to 81.0, LiveCodeBench v6 63.5 to 73.3, fewer hallucinations and Tau-bench 53.5 (airline) / 63.9 (retail) with function calling.

• [4]70%
  The 'minor upgrade' that's anything but: DeepSeek R1-0528 deep dive - RunPod

  runpod.io· Added Oct 12, 2026· 19% of score

  Independent analysis written the same week: the reasoning-depth change (about 12K to 23K tokens per AIME question), AIME 2025 87.5% against o3's 88.9% and Gemini 2.5 Pro's 83.0%, LiveCodeBench 73.1 and the distilled 8B model.^[1]

• [5]75%
  DeepSeek R1 0528 - OpenRouter endpoints

  openrouter.ai· Added Oct 12, 2026· 19% of score

  OpenRouter's record: created 2025-05-28, 671B parameters with 37B active, 163,840-token context and provider prices of $0.50 to $0.70 per million input tokens and $2.15 to $2.50 output.

Suggest a correction

Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.