Speech and Language Processing (3rd Edition Draft)
by Dan Jurafsky & James H. Martin (Stanford)
The canonical NLP textbook, rewritten around transformers, LLMs, and modern language modeling.
Overview
Speech and Language Processing by Dan Jurafsky and James H. Martin is the standard graduate/undergraduate NLP textbook, and the third edition (a continually revised draft, with the online manuscript most recently updated on January 6, 2026) rebuilds the material around large language models. It is offered free as individual chapter PDFs plus a full compiled PDF and lecture slides. Volume I ('Large Language Models') covers words and tokenization, n-gram models, embeddings, neural networks, transformers, large language models, post-training and alignment, masked language models, retrieval-augmented generation, machine translation, RNNs/LSTMs, phonetics, automatic speech recognition, and text-to-speech. Volume II ('Linguistic Structure') covers sequence labeling, constituency and dependency parsing, information extraction, semantic role labeling, sentiment analysis, coreference resolution, and discourse. The authors explicitly encourage classroom use and welcome reader feedback.
At a Glance
- Topic
- Models
- Level
- Intermediate
- Format
- Book
- Cost
- Free
- Duration
- Full textbook, self-paced
- Provider
- Dan Jurafsky & James H. Martin (Stanford)
- Hands-on
- No
- Certificate
- None
What You’ll Learn
- ✓How transformers and large language models work from tokenization to attention
- ✓Core NLP foundations: embeddings, n-gram and neural language models, sequence labeling
- ✓Post-training and alignment, plus retrieval-augmented generation (RAG) fundamentals
- ✓Machine translation, automatic speech recognition, and text-to-speech pipelines
- ✓Linguistic structure: parsing, information extraction, coreference, and discourse
Highlights
- •The most widely assigned NLP textbook, updated for the LLM era
- •Completely free: chapter PDFs, a full compiled PDF, and teaching slides
- •Draft actively revised, with the latest online manuscript dated January 6, 2026
Who It’s For
Best For
- ✓Engineers who want the theory beneath the models they build with
- ✓Students taking or self-studying an NLP / LLM course
- ✓Anyone needing an authoritative reference on transformers and language modeling
Prerequisites
- •Basic probability, linear algebra, and calculus
- •Comfort with programming and introductory machine learning
FAQ
What is Speech and Language Processing (3rd Edition Draft)?
The free online draft of Jurafsky & Martin's classic NLP textbook, extensively updated for the LLM era. It is the foundational reference for students and engineers who want to understand how modern language models work from tokens up through transformers, alignment, and RAG.
Is Speech and Language Processing (3rd Edition Draft) free?
Speech and Language Processing (3rd Edition Draft) is free to access.
What level is Speech and Language Processing (3rd Edition Draft) for?
Speech and Language Processing (3rd Edition Draft) is aimed at a intermediate audience. Recommended background: Basic probability, linear algebra, and calculus, Comfort with programming and introductory machine learning.
How long does Speech and Language Processing (3rd Edition Draft) take?
Expect roughly Full textbook, self-paced. Most learners work through it at their own pace.
What will I learn from Speech and Language Processing (3rd Edition Draft)?
You'll learn: How transformers and large language models work from tokenization to attention; Core NLP foundations: embeddings, n-gram and neural language models, sequence labeling; Post-training and alignment, plus retrieval-augmented generation (RAG) fundamentals; Machine translation, automatic speech recognition, and text-to-speech pipelines; Linguistic structure: parsing, information extraction, coreference, and discourse.