stanfordnlp · TomeHirata · Nov 13, 2025 · Nov 13, 2025 · Nov 13, 2025
diff --git a/docs/docs/cheatsheet.md b/docs/docs/cheatsheet.md
@@ -67,7 +67,7 @@ print(f"Final Predicted Answer (after ReAct process): {result.answer}")
 
 ```python
 colbertv2_wiki17_abstracts = dspy.ColBERTv2(url='http://20.102.90.50:2017/wiki17_abstracts')
-dspy.settings.configure(rm=colbertv2_wiki17_abstracts)
+dspy.configure(rm=colbertv2_wiki17_abstracts)
 
 #Define Retrieve Module
 retriever = dspy.Retrieve(k=3)
@@ -450,7 +450,7 @@ asyncio.run(dspy_program(question="What is DSPy"))
 
 ```python
 import dspy
-dspy.settings.configure(track_usage=True)
+dspy.configure(track_usage=True)
 
 result = dspy.ChainOfThought(BasicQA)(question="What is 2+2?")
 print(f"Token usage: {result.get_lm_usage()}")

diff --git a/docs/docs/learn/programming/language_models.md b/docs/docs/learn/programming/language_models.md
@@ -229,7 +229,7 @@ To enable the Responses API, just set `model_type="responses"` when creating the
 import dspy
 
 # Configure DSPy to use the Responses API for your language model
-dspy.settings.configure(
+dspy.configure(
     lm=dspy.LM(
         "openai/gpt-5-mini",
         model_type="responses",

diff --git a/docs/docs/learn/programming/modules.md b/docs/docs/learn/programming/modules.md
@@ -260,7 +260,7 @@ print(hop(claim="Stephen Curry is the best 3 pointer shooter ever in the human h
 DSPy provides built-in tracking of language model usage across all module calls. To enable tracking:
 
 ```python
-dspy.settings.configure(track_usage=True)
+dspy.configure(track_usage=True)
 ```
 
 Once enabled, you can access usage statistics from any `dspy.Prediction` object:
@@ -275,7 +275,7 @@ The usage data is returned as a dictionary that maps each language model name to
 import dspy
 
 # Configure DSPy with tracking enabled
-dspy.settings.configure(
+dspy.configure(
     lm=dspy.LM("openai/gpt-4o-mini", cache=False),
     track_usage=True
 )
@@ -326,7 +326,7 @@ When using DSPy's caching features (either in-memory or on-disk via litellm), ca
 
 ```python
 # Enable caching
-dspy.settings.configure(
+dspy.configure(
     lm=dspy.LM("openai/gpt-4o-mini", cache=True),
     track_usage=True
 )

diff --git a/docs/docs/tutorials/audio/index.ipynb b/docs/docs/tutorials/audio/index.ipynb
@@ -117,7 +117,7 @@
     "Now let's configure our LLM which can process input audio. \n",
     "\n",
     "```python\n",
-    "dspy.settings.configure(lm=dspy.LM(model='gpt-4o-mini-audio-preview-2024-12-17'))\n",
+    "dspy.configure(lm=dspy.LM(model='gpt-4o-mini-audio-preview-2024-12-17'))\n",
     "```\n",
     "\n",
     "Note: Using `dspy.Audio` in signatures allows passing in audio directly to the model. "
@@ -332,7 +332,7 @@
     "        audio = generate_dspy_audio(raw_line, out.openai_instruction)\n",
     "        return dspy.Prediction(audio=audio)\n",
     "    \n",
-    "dspy.settings.configure(lm=dspy.LM(model='gpt-4o-mini'))"
+    "dspy.configure(lm=dspy.LM(model='gpt-4o-mini'))"
    ]
   },
   {

diff --git a/docs/docs/tutorials/cache/index.md b/docs/docs/tutorials/cache/index.md
@@ -25,7 +25,7 @@ import time
 
 os.environ["OPENAI_API_KEY"] = "{your_openai_key}"
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"), track_usage=True)
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"), track_usage=True)
 
 predict = dspy.Predict("question->answer")
 
@@ -167,7 +167,7 @@ import time
 
 os.environ["OPENAI_API_KEY"] = "{your_openai_key}"
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
 predict = dspy.Predict("question->answer")
 
@@ -193,7 +193,7 @@ from hashlib import sha256
 
 os.environ["OPENAI_API_KEY"] = "{your_openai_key}"
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
 class CustomCache(dspy.clients.Cache):
 

diff --git a/docs/docs/tutorials/conversation_history/index.md b/docs/docs/tutorials/conversation_history/index.md
@@ -12,7 +12,7 @@ import os
 
 os.environ["OPENAI_API_KEY"] = "{your_openai_api_key}"
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
 class QA(dspy.Signature):
     question: str = dspy.InputField()
@@ -121,7 +121,7 @@ For example:
 ```python
 import dspy
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
 
 class QA(dspy.Signature):

diff --git a/docs/docs/tutorials/deployment/index.md b/docs/docs/tutorials/deployment/index.md
@@ -7,7 +7,7 @@ Below, we'll assume you have the following simple DSPy program that you want to
 ```python
 import dspy
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 dspy_program = dspy.ChainOfThought("question -> answer")
 ```
 
@@ -40,7 +40,7 @@ class Question(BaseModel):
 
 # Configure your language model and 'asyncify' your DSPy program.
 lm = dspy.LM("openai/gpt-4o-mini")
-dspy.settings.configure(lm=lm, async_max_workers=4) # default is 8
+dspy.configure(lm=lm, async_max_workers=4) # default is 8
 dspy_program = dspy.ChainOfThought("question -> answer")
 dspy_program = dspy.asyncify(dspy_program)
 
@@ -163,7 +163,7 @@ mlflow.set_tracking_uri("http://127.0.0.1:5000/")
 mlflow.set_experiment("deploy_dspy_program")
 
 lm = dspy.LM("openai/gpt-4o-mini")
-dspy.settings.configure(lm=lm)
+dspy.configure(lm=lm)
 
 class MyProgram(dspy.Module):
     def __init__(self):

diff --git a/docs/docs/tutorials/entity_extraction/index.ipynb b/docs/docs/tutorials/entity_extraction/index.ipynb
@@ -172,7 +172,7 @@
         "Specifically, we'll:\n",
         "- Create a `PeopleExtraction` DSPy Signature to specify the input (`tokens`) and output (`extracted_people`) fields.\n",
         "- Define a `people_extractor` program that uses DSPy's built-in `dspy.ChainOfThought` module to implement the `PeopleExtraction` signature. The program extracts entities referring to people from a list of input tokens using language model (LM) prompting.\n",
-        "- Use the `dspy.LM` class and `dspy.settings.configure()` method to configure the language model that DSPy will use when invoking the program."
+        "- Use the `dspy.LM` class and `dspy.configure()` method to configure the language model that DSPy will use when invoking the program."
       ]
     },
     {
@@ -208,7 +208,7 @@
       "outputs": [],
       "source": [
         "lm = dspy.LM(model=\"openai/gpt-4o-mini\")\n",
-        "dspy.settings.configure(lm=lm)"
+        "dspy.configure(lm=lm)"
       ]
     },
     {

diff --git a/docs/docs/tutorials/image_generation_prompting/index.ipynb b/docs/docs/tutorials/image_generation_prompting/index.ipynb
@@ -80,7 +80,7 @@
     "from IPython.display import display\n",
     "\n",
     "lm = dspy.LM(model=\"gpt-4o-mini\", temperature=0.5)\n",
-    "dspy.settings.configure(lm=lm)"
+    "dspy.configure(lm=lm)"
    ]
   },
   {

diff --git a/docs/docs/tutorials/program_of_thought/index.ipynb b/docs/docs/tutorials/program_of_thought/index.ipynb
@@ -69,7 +69,7 @@
    "source": [
     "llama31_70b = dspy.LM(\"openai/meta-llama/Meta-Llama-3-70b-Instruct\", api_base=\"API_BASE\", api_key=\"None\")\n",
     "\n",
-    "dspy.settings.configure(lm=llama31_70b)"
+    "dspy.configure(lm=llama31_70b)"
    ]
   },
   {

diff --git a/docs/docs/tutorials/saving/index.md b/docs/docs/tutorials/saving/index.md
@@ -20,7 +20,7 @@ Let's say we have compiled a program with some data, and we want to save the pro
 import dspy
 from dspy.datasets.gsm8k import GSM8K, gsm8k_metric
 
-dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
 gsm8k = GSM8K()
 gsm8k_trainset = gsm8k.train[:10]

diff --git a/docs/docs/tutorials/streaming/index.md b/docs/docs/tutorials/streaming/index.md
@@ -130,7 +130,7 @@ import asyncio
 import dspy
 
 lm = dspy.LM("openai/gpt-4o-mini", cache=False)
-dspy.settings.configure(lm=lm)
+dspy.configure(lm=lm)
 
 
 class MyModule(dspy.Module):
@@ -204,7 +204,7 @@ import asyncio
 import dspy
 
 lm = dspy.LM("openai/gpt-4o-mini", cache=False)
-dspy.settings.configure(lm=lm)
+dspy.configure(lm=lm)
 
 
 def fetch_user_info(user_name: str):
@@ -259,7 +259,7 @@ import asyncio
 import dspy
 
 lm = dspy.LM("openai/gpt-4o-mini", cache=False)
-dspy.settings.configure(lm=lm)
+dspy.configure(lm=lm)
 
 
 class MyModule(dspy.Module):
@@ -371,7 +371,7 @@ import asyncio
 import dspy
 
 lm = dspy.LM("openai/gpt-4o-mini", cache=False)
-dspy.settings.configure(lm=lm)
+dspy.configure(lm=lm)
 
 
 class MyModule(dspy.Module):

diff --git a/dspy/adapters/types/history.py b/dspy/adapters/types/history.py
@@ -22,7 +22,7 @@ class MySignature(dspy.Signature):
         ```
         import dspy
 
-        dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+        dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
         class MySignature(dspy.Signature):
             question: str = dspy.InputField()
@@ -44,7 +44,7 @@ class MySignature(dspy.Signature):
         ```
         import dspy
 
-        dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+        dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
         class MySignature(dspy.Signature):
             question: str = dspy.InputField()

diff --git a/dspy/datasets/dataloader.py b/dspy/datasets/dataloader.py
@@ -132,7 +132,7 @@ def from_rm(self, num_samples: int, fields: list[str], input_keys: list[str]) ->
                 )
         except AttributeError:
             raise ValueError(
-                "Retrieval module not found. Please set a retrieval module using `dspy.settings.configure`."
+                "Retrieval module not found. Please set a retrieval module using `dspy.configure`."
             )
 
     def sample(

diff --git a/dspy/dsp/utils/settings.py b/dspy/dsp/utils/settings.py
@@ -151,7 +151,7 @@ def _ensure_configure_allowed(self):
 
         if not in_ipython and config_owner_async_task != asyncio.current_task():
             raise RuntimeError(
-                "dspy.settings.configure(...) can only be called from the same async task that called it first. Please "
+                "dspy.configure(...) can only be called from the same async task that called it first. Please "
                 "use `dspy.context(...)` in other async tasks instead."
             )
 

diff --git a/dspy/predict/best_of_n.py b/dspy/predict/best_of_n.py
@@ -29,7 +29,7 @@ def __init__(
             ```python
             import dspy
 
-            dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+            dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
             # Define a QA module with chain of thought
             qa = dspy.ChainOfThought("question -> answer")

diff --git a/dspy/predict/refine.py b/dspy/predict/refine.py
@@ -67,7 +67,7 @@ def __init__(
             ```python
             import dspy
 
-            dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+            dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
             # Define a QA module with chain of thought
             qa = dspy.ChainOfThought("question -> answer")

diff --git a/dspy/propose/dataset_summary_generator.py b/dspy/propose/dataset_summary_generator.py
@@ -50,7 +50,7 @@ def create_dataset_summary(trainset, view_data_batch_size, prompt_model, log_fil
         print("\nBootstrapping dataset summary (this will be used to generate instructions)...")
     upper_lim = min(len(trainset), view_data_batch_size)
     prompt_model = prompt_model if prompt_model else dspy.settings.lm
-    with dspy.settings.context(lm=prompt_model):
+    with dspy.context(lm=prompt_model):
         observation = dspy.Predict(DatasetDescriptor, n=1, temperature=1.0)(examples=order_input_keys_in_string(trainset[0:upper_lim].__repr__()))
     observations = observation["observations"]
 
@@ -68,7 +68,7 @@ def create_dataset_summary(trainset, view_data_batch_size, prompt_model, log_fil
             if verbose:
                 print(f"b: {b}")
             upper_lim = min(len(trainset), b+view_data_batch_size)
-            with dspy.settings.context(lm=prompt_model):
+            with dspy.context(lm=prompt_model):
                 output = dspy.Predict(DatasetDescriptorWithPriorObservations, n=1, temperature=1.0)(prior_observations=observations, examples=order_input_keys_in_string(trainset[b:upper_lim].__repr__()))
             if len(output["observations"]) >= 8 and output["observations"][:8].upper() == "COMPLETE":
                 skips += 1
@@ -84,7 +84,7 @@ def create_dataset_summary(trainset, view_data_batch_size, prompt_model, log_fil
             print(f"e {e}. using observations from past round for a summary.")
 
     if prompt_model:
-        with dspy.settings.context(lm=prompt_model):
+        with dspy.context(lm=prompt_model):
             summary = dspy.Predict(ObservationSummarizer, n=1, temperature=1.0)(observations=observations)
     else:
         summary = dspy.Predict(ObservationSummarizer, n=1, temperature=1.0)(observations=observations)

diff --git a/dspy/propose/grounded_proposer.py b/dspy/propose/grounded_proposer.py
@@ -419,7 +419,7 @@ def propose_instruction_for_predictor(
             temperature=self.init_temperature,
         )
 
-        with dspy.settings.context(lm=rollout_lm):
+        with dspy.context(lm=rollout_lm):
             proposed_instruction = instruction_generator(
                 demo_candidates=demo_candidates,
                 pred_i=pred_i,

diff --git a/dspy/retrievers/weaviate_rm.py b/dspy/retrievers/weaviate_rm.py
@@ -34,7 +34,7 @@ class WeaviateRM(dspy.Retrieve):
         llm = dspy.Cohere(model="command-r-plus", api_key=api_key)
         weaviate_client = weaviate.connect_to_[local, wcs, custom, embedded]("your-path-here")
         retriever_model = WeaviateRM("my_collection_name", weaviate_client=weaviate_client)
-        dspy.settings.configure(lm=llm, rm=retriever_model)
+        dspy.configure(lm=llm, rm=retriever_model)
 
         retrieve = dspy.Retrieve(k=1)
         topK_passages = retrieve("what are the stages in planning, sanctioning and execution of public works").passages

diff --git a/dspy/streaming/streamify.py b/dspy/streaming/streamify.py
@@ -64,7 +64,7 @@ def streamify(
     import asyncio
     import dspy
 
-    dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+    dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
     # Create the program and wrap it with streaming functionality
     program = dspy.streamify(dspy.Predict("q->a"))
 
@@ -88,7 +88,7 @@ async def use_streaming():
     import asyncio
     import dspy
 
-    dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini"))
+    dspy.configure(lm=dspy.LM("openai/gpt-4o-mini"))
 
     class MyStatusMessageProvider(StatusMessageProvider):
         def module_start_status_message(self, instance, inputs):
@@ -121,7 +121,7 @@ async def use_streaming():
     import asyncio
     import dspy
 
-    dspy.settings.configure(lm=dspy.LM("openai/gpt-4o-mini", cache=False))
+    dspy.configure(lm=dspy.LM("openai/gpt-4o-mini", cache=False))
 
     # Create the program and wrap it with streaming functionality
     predict = dspy.Predict("question->answer, reasoning")

diff --git a/dspy/teleprompt/bootstrap.py b/dspy/teleprompt/bootstrap.py
@@ -182,13 +182,13 @@ def _bootstrap_one_example(self, example, round_idx=0):
         predictor_cache = {}
 
         try:
-            with dspy.settings.context(trace=[], **self.teacher_settings):
+            with dspy.context(trace=[], **self.teacher_settings):
                 lm = dspy.settings.lm
                 # Use a fresh rollout with temperature=1.0 to bypass caches.
                 lm = lm.copy(rollout_id=round_idx, temperature=1.0) if round_idx > 0 else lm
                 new_settings = {"lm": lm} if round_idx > 0 else {}
 
-                with dspy.settings.context(**new_settings):
+                with dspy.context(**new_settings):
                     for name, predictor in teacher.named_predictors():
                         predictor_cache[name] = predictor.demos
                         predictor.demos = [x for x in predictor.demos if x != example]

diff --git a/dspy/teleprompt/copro_optimizer.py b/dspy/teleprompt/copro_optimizer.py
@@ -20,7 +20,7 @@
 
 Note that this teleprompter takes in the following parameters:
 
-* prompt_model: The model used for prompt generation. When unspecified, defaults to the model set in settings (ie. dspy.settings.configure(lm=task_model)).
+* prompt_model: The model used for prompt generation. When unspecified, defaults to the model set in settings (ie. dspy.configure(lm=task_model)).
 * metric: The task metric used for optimization.
 * breadth: The number of new prompts to generate at each iteration. Default=10.
 * depth: The number of times we should ask our prompt model to generate new prompts, with the history of the past prompts as input. Default=3.
@@ -156,7 +156,7 @@ def compile(self, student, *, trainset, eval_kwargs):
             basic_instruction = self._get_signature(predictor).instructions
             basic_prefix = self._get_signature(predictor).fields[last_key].json_schema_extra["prefix"]
             if self.prompt_model:
-                with dspy.settings.context(lm=self.prompt_model):
+                with dspy.context(lm=self.prompt_model):
                     instruct = dspy.Predict(
                         BasicGenerateInstruction,
                         n=self.breadth - 1,
@@ -306,7 +306,7 @@ def compile(self, student, *, trainset, eval_kwargs):
 
                 # Generate next batch of potential prompts to optimize, with previous attempts as input
                 if self.prompt_model:
-                    with dspy.settings.context(lm=self.prompt_model):
+                    with dspy.context(lm=self.prompt_model):
                         instr = dspy.Predict(
                             GenerateInstructionGivenAttempts,
                             n=self.breadth,