NLP Lineage

22 May 2025 · Reasoning

Claude 4

Claude 4 (Opus 4 and Sonnet 4)

Anthropic

Date
22 May 2025
What the date marks
Model release
Kind
Model
Organisation
Anthropic
Access
API only
Threads
In the guided tour
Chapter 13. Models that act
Learning phases
7. Agents

A milestone: 1 later milestone descends directly from it, and its descendants reach 2 of the 10 threads. How levels are decided

Figures quoted here are reported by the organisation that built it, not independently verified.

Read the primary source (anthropic.com)

Full summary in progress. Until then, this page shows the milestone's record from the research data.

What it changed

Before it

Coding agents worked in short bursts and lost the thread on longer tasks.

The problem

Real engineering tasks run for hours and require holding context across many steps.

What it introduced

Hybrid reasoning models tuned for sustained agentic work, reported at 72.5% on SWE-bench Verified.

Why it matters

It set the agentic coding bar for the year and made multi-hour autonomous tasks a selling point. The benchmark figures are vendor-reported.

What modern models inherit

Long-horizon agentic work as the thing frontier models compete on.

What came next

The 4.5 line in late 2025, then Claude 5 in 2026.

Before and after

The milestones it descends from, and the milestones that descend from it. Work developed alongside it is not counted. Weak links are left out unless you include them.

Before: where it came from

Claude 4 has 47 documented ancestors across 10 chapters, reaching back to Recurrent networks in 1990. Showing its direct parents and the 18 turning points among them.

Claude 4 has 47 documented ancestors across 10 chapters, reaching back to Recurrent networks in 1990. Showing all of them.

Claude 4 has 51 recorded ancestors, including indirect ones, across 10 chapters, reaching back to Recurrent networks in 1990. Showing its direct parents and the 18 turning points among them.

Claude 4 has 51 recorded ancestors, including indirect ones, across 10 chapters, reaching back to Recurrent networks in 1990. Showing all of them.

  1. 2025Claude 4This milestone.
  2. 2025Claude 3.7Claude 4 built directly on thisChapter 12. Thinking longer
  3. 2024o1 Turning pointClaude 3.7 was influenced by thisChapter 12. Thinking longer
  4. 2024Claude 3.5 SonnetClaude 3.7 built directly on thisChapter 14. Where the threads meet now
  5. 2024Claude 3Claude 4 succeeded thisChapter 14. Where the threads meet now
  6. 2023SWE-benchClaude 4 was influenced by thisChapter 13. Models that act
  7. 2023GPT-4 Turning pointSWE-bench was influenced by this
    Claude 3 was influenced by this    Chapter 7. The assistant goes public
  8. 2023ClaudeClaude 3 built directly on thisChapter 7. The assistant goes public
  9. 2022Constitutional AIClaude 3 built directly on thisChapter 6. Teaching models what people want
  10. 2022GPT-3.5 Turning pointGPT-4 succeeded thisChapter 7. The assistant goes public
  11. 2022HH-RLHFConstitutional AI built directly on this
    Claude was influenced by this    Chapter 6. Teaching models what people want
  12. 2022InstructGPT Turning pointGPT-4 built directly on thisChapter 6. Teaching models what people want
  13. 2022Chain of thought Turning pointo1 was influenced by this
    Constitutional AI was influenced by this    Chapter 12. Thinking longer
  14. 2021HHH assistantHH-RLHF built directly on thisChapter 6. Teaching models what people want
  15. 2021FLANInstructGPT was influenced by thisChapter 6. Teaching models what people want
  16. 2021Codex / HumanEvalSWE-bench built directly on thisChapter 13. Models that act
  17. 2020RLHF for summarisation Turning pointInstructGPT built directly on this
    HH-RLHF built directly on this    Chapter 6. Teaching models what people want
  18. 2020GPT-3 Turning pointCodex / HumanEval built directly on thisChapter 5. Scale becomes a science
  19. 2020Scaling laws Turning pointGPT-4 built directly on thisChapter 5. Scale becomes a science
  20. 2019T5 Turning pointFLAN was influenced by thisChapter 4. Pretrain once, use everywhere
  21. 2019C4 Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  22. 2019RLHF for LMsRLHF for summarisation built directly on thisChapter 6. Teaching models what people want
  23. 2019Sparse Transformer Turning pointGPT-3 built directly on thisChapter 4. Pretrain once, use everywhere
  24. 2019GPT-2 Turning pointScaling laws was influenced by this
    GPT-3 built directly on this    Chapter 4. Pretrain once, use everywhere
  25. 2019WebTextGPT-2 was influenced by thisChapter 4. Pretrain once, use everywhere
  26. 2018BERT Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  27. 2018SentencePieceT5 built directly on thisChapter 4. Pretrain once, use everywhere
  28. 2018GPT-1GPT-2 built directly on thisChapter 4. Pretrain once, use everywhere
  29. 2018ELMoBERT was influenced by thisChapter 4. Pretrain once, use everywhere
  30. 2018ULMFiTBERT was influenced by thisChapter 4. Pretrain once, use everywhere
  31. 2017CoVeELMo was influenced by thisChapter 3. Attention, then only attention
  32. 2017RLHF originConstitutional AI was influenced by thisChapter 6. Teaching models what people want
  33. 2017Transformer Turning pointScaling laws built directly on thisChapter 3. Attention, then only attention
  34. 2016Layer normTransformer built directly on this
    GPT-2 built directly on this    Chapter 3. Attention, then only attention
  35. 2015Residual connectionsTransformer was influenced by thisChapter 3. Attention, then only attention
  36. 2015LM pretraining Turning pointGPT-1 was influenced by thisChapter 2. Words become vectors
  37. 2015BPE tokenizationGPT-2 built directly on thisChapter 2. Words become vectors
  38. 2015Luong attentionTransformer was influenced by thisChapter 3. Attention, then only attention
  39. 2014AdamTransformer was influenced by thisChapter 2. Words become vectors
  40. 2014GloVeELMo built directly on thisChapter 2. Words become vectors
  41. 2014Seq2seq Turning pointTransformer built directly on thisChapter 2. Words become vectors
  42. 2014Attention Turning pointTransformer built directly on thisChapter 3. Attention, then only attention
  43. 2014GRU / encoder-decoderAttention built directly on thisChapter 2. Words become vectors
  44. 2013word2vecELMo was influenced by thisChapter 2. Words become vectors
  45. 2011NLP from scratchELMo was influenced by thisChapter 1. Language as probability
  46. 2003Neural LMNLP from scratch was influenced by this
    word2vec was influenced by this    Chapter 1. Language as probability
  47. 1997LSTM Turning pointSeq2seq built directly on thisChapter 1. Language as probability
  48. 1990Recurrent networksLSTM was influenced by thisChapter 1. Language as probability
  1. 2025Claude 4This milestone.
  2. 2025Claude 3.7Claude 4 built directly on thisChapter 12. Thinking longer
  3. 2024o1 Turning pointClaude 3.7 was influenced by thisChapter 12. Thinking longer
  4. 2024Claude 3.5 SonnetClaude 3.7 built directly on thisChapter 14. Where the threads meet now
  5. 2024Claude 3Claude 4 succeeded thisChapter 14. Where the threads meet now
  6. 2023SWE-benchClaude 4 was influenced by thisChapter 13. Models that act
  7. 2023Process supervisiono1 was influenced by this (indirect influence)Chapter 12. Thinking longer
  8. 2023GPT-4 Turning pointSWE-bench was influenced by this
    Claude 3 was influenced by this    Chapter 7. The assistant goes public
  9. 2023ClaudeClaude 3 built directly on thisChapter 7. The assistant goes public
  10. 2022Constitutional AIClaude 3 built directly on thisChapter 6. Teaching models what people want
  11. 2022GPT-3.5 Turning pointGPT-4 succeeded thisChapter 7. The assistant goes public
  12. 2022HH-RLHFConstitutional AI built directly on this
    Claude was influenced by this    Chapter 6. Teaching models what people want
  13. 2022STaRo1 was influenced by this (indirect influence)Chapter 12. Thinking longer
  14. 2022InstructGPT Turning pointGPT-4 built directly on thisChapter 6. Teaching models what people want
  15. 2022Chain of thought Turning pointo1 was influenced by this
    Constitutional AI was influenced by this    Chapter 12. Thinking longer
  16. 2021HHH assistantHH-RLHF built directly on thisChapter 6. Teaching models what people want
  17. 2021FLANInstructGPT was influenced by thisChapter 6. Teaching models what people want
  18. 2021Codex / HumanEvalSWE-bench built directly on thisChapter 13. Models that act
  19. 2020RLHF for summarisation Turning pointInstructGPT built directly on this
    HH-RLHF built directly on this    Chapter 6. Teaching models what people want
  20. 2020GPT-3 Turning pointCodex / HumanEval built directly on thisChapter 5. Scale becomes a science
  21. 2020Scaling laws Turning pointGPT-4 built directly on thisChapter 5. Scale becomes a science
  22. 2019T5 Turning pointFLAN was influenced by thisChapter 4. Pretrain once, use everywhere
  23. 2019C4 Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  24. 2019RLHF for LMsRLHF for summarisation built directly on thisChapter 6. Teaching models what people want
  25. 2019Sparse Transformer Turning pointGPT-3 built directly on thisChapter 4. Pretrain once, use everywhere
  26. 2019GPT-2 Turning pointScaling laws was influenced by this
    GPT-3 built directly on this    Chapter 4. Pretrain once, use everywhere
  27. 2019WebTextGPT-2 was influenced by thisChapter 4. Pretrain once, use everywhere
  28. 2019Transformer-XLSparse Transformer was influenced by this (indirect influence)Chapter 4. Pretrain once, use everywhere
  29. 2018BERT Turning pointT5 was influenced by thisChapter 4. Pretrain once, use everywhere
  30. 2018SentencePieceT5 built directly on thisChapter 4. Pretrain once, use everywhere
  31. 2018GPT-1GPT-2 built directly on thisChapter 4. Pretrain once, use everywhere
  32. 2018ELMoGPT-1 was influenced by this (indirect influence)Chapter 4. Pretrain once, use everywhere
  33. 2018ULMFiTGPT-1 was influenced by this (indirect influence)Chapter 4. Pretrain once, use everywhere
  34. 2017CoVeELMo was influenced by thisChapter 3. Attention, then only attention
  35. 2017RLHF origino1 was influenced by this (indirect influence)
    Constitutional AI was influenced by this    Chapter 6. Teaching models what people want
  36. 2017Transformer Turning pointScaling laws built directly on thisChapter 3. Attention, then only attention
  37. 2017ConvS2STransformer was influenced by this (indirect influence)Chapter 3. Attention, then only attention
  38. 2016Layer normTransformer built directly on this
    GPT-2 built directly on this    Chapter 3. Attention, then only attention
  39. 2015Residual connectionsTransformer was influenced by thisChapter 3. Attention, then only attention
  40. 2015LM pretraining Turning pointGPT-1 was influenced by thisChapter 2. Words become vectors
  41. 2015BPE tokenizationGPT-2 built directly on thisChapter 2. Words become vectors
  42. 2015Luong attentionTransformer was influenced by thisChapter 3. Attention, then only attention
  43. 2014AdamTransformer was influenced by thisChapter 2. Words become vectors
  44. 2014GloVeELMo built directly on thisChapter 2. Words become vectors
  45. 2014Seq2seq Turning pointTransformer built directly on thisChapter 2. Words become vectors
  46. 2014Attention Turning pointTransformer built directly on thisChapter 3. Attention, then only attention
  47. 2014GRU / encoder-decoderAttention built directly on thisChapter 2. Words become vectors
  48. 2013word2vecELMo was influenced by thisChapter 2. Words become vectors
  49. 2011NLP from scratchELMo was influenced by thisChapter 1. Language as probability
  50. 2003Neural LMNLP from scratch was influenced by this
    word2vec was influenced by this    Chapter 1. Language as probability
  51. 1997LSTM Turning pointSeq2seq built directly on thisChapter 1. Language as probability
  52. 1990Recurrent networksLSTM was influenced by thisChapter 1. Language as probability

After: what it led to

Claude 4 has 8 documented descendants across 2 chapters, reaching forward to Claude Opus 5.5 in 2026. Showing its direct descendants.

Claude 4 has 8 documented descendants across 2 chapters, reaching forward to Claude Opus 5.5 in 2026. Showing all of them.

Claude 4 has 8 recorded descendants, including indirect ones, across 2 chapters, reaching forward to Claude Opus 5.5 in 2026. Showing its direct descendants.

Claude 4 has 8 recorded descendants, including indirect ones, across 2 chapters, reaching forward to Claude Opus 5.5 in 2026. Showing all of them.

  1. 2025Claude 4This milestone.
  2. 2025Claude Sonnet 4.5built directly on Claude 4Chapter 13. Models that act
  3. 2025Claude Opus 4.5built directly on Claude Sonnet 4.5Chapter 13. Models that act
  4. 2026Claude Opus 4.6built directly on Claude Opus 4.5Chapter 14. Where the threads meet now
  5. 2026Claude Opus 4.7built directly on Claude Opus 4.6Chapter 14. Where the threads meet now
  6. 2026Claude Opus 4.8built directly on Claude Opus 4.7Chapter 14. Where the threads meet now
  7. 2026Claude Sonnet 5succeeded Claude Opus 4.8Chapter 14. Where the threads meet now
  8. 2026Claude Opus 5succeeded Claude Opus 4.8Chapter 14. Where the threads meet now
  9. 2026Claude Opus 5.5built directly on Claude Opus 5Chapter 14. Where the threads meet now
  1. 2025Claude 4This milestone.
  2. 2025Claude Sonnet 4.5built directly on Claude 4Chapter 13. Models that act
  3. 2025Claude Opus 4.5built directly on Claude Sonnet 4.5Chapter 13. Models that act
  4. 2026Claude Opus 4.6built directly on Claude Opus 4.5Chapter 14. Where the threads meet now
  5. 2026Claude Opus 4.7built directly on Claude Opus 4.6Chapter 14. Where the threads meet now
  6. 2026Claude Opus 4.8built directly on Claude Opus 4.7Chapter 14. Where the threads meet now
  7. 2026Claude Sonnet 5succeeded Claude Opus 4.8Chapter 14. Where the threads meet now
  8. 2026Claude Opus 5succeeded Claude Opus 4.8Chapter 14. Where the threads meet now
  9. 2026Claude Opus 5.5built directly on Claude Opus 5Chapter 14. Where the threads meet now

Direct links and their sources

Where it came from

Built directly on

Influenced by

Succeeded

What it led to

Built on directly by