Skip to content
LMQL

LMQLLMQL is a programming language for LLMs.

Robust and modular LLM prompting using types, templates, constraints and an optimizing runtime.

lmql
@lmql.query def meaning_of_life(): '''lmql # top-level strings are prompts "Q: What is the answer to life, the \ universe and everything?" # generation via (constrained) variables "A: [ANSWER]" where \ len(ANSWER) < 120 and STOPS_AT(ANSWER, ".") # results are directly accessible print("LLM returned", ANSWER) # use typed variables for guaranteed # output format "The answer is [NUM: int]" # query programs are just functions return NUM ''' # so from Python, you can just do this meaning_of_life() # 42

Created by the SRI Lab @ ETH Zurich and contributors.


Nested Queries bring Procedural Programming to Prompting NEW

LMQL now supports nested queries, enabling modularized local instructions and re-use of prompt components.


Learn more

promptdown

Execution Trace

Q: When was Obama born?200incontext

200ANSWER04/08/1961200incontext200incontext200 Q: When was Bruno Mars born?200incontext1200ANSWER08/10/1985200incontext1200incontext1200 Q: When was Dua Lipa born?200incontext2200ANSWER22/08/1995200incontext2200incontext2200 Out of these, who was born last?LASTDua Lipa

Works Across Backends

LMQL automatically makes your LLM code portable across several backends. You can switch between them with a single line of code.

Explore LMQL

Prompt construction and generation is implemented via expressive Python control flow and string interpolation.

LMQL

lmql
# top level strings are prompts
"My packing list for the trip:"

# use loops for repeated prompts
for i in range(4):
    # 'where' denotes hard constraints enforced by the runtime
    "- [THING] \n" where THING in \ 
        ["Volleyball", "Sunscreen", "Bathing Suit"]

Model Output

promptdown

My packing list for the trip: - THING Volleyball - THING Bathing Suit - THING Sunscreen - THING Volleyball