dspy-lite

DSPy with litellm replaced by litelm and numpy made optional.

What this is

Same DSPy 3.2.1 -- signatures, modules (Predict, CoT, ReAct), optimizers (Bootstrap, MIPRO, SIMBA, GEPA, GRPO)
litellm (100k+ LOC, 70+ transitive deps) replaced by litelm (2.6k LOC, 2 deps)
numpy moved from core dep to optional [embeddings] extra
83% smaller clean install (24 MB vs 144 MB, 36 packages vs 67)
Unaffected by the litellm supply chain compromise (CVE-2026-33634) — zero litellm code or dependencies

Install

pip install dspy-lite

Supported Python versions: 3.10 through 3.14.

With provider extras:

pip install dspy-lite[anthropic]     # Anthropic
pip install dspy-lite[bedrock]       # AWS Bedrock
pip install dspy-lite[embeddings]    # numpy for embeddings
pip install dspy-lite[all]           # everything

Usage

import dspy

lm = dspy.LM("openai/gpt-4o-mini")
dspy.configure(lm=lm)

predict = dspy.Predict("question -> answer")
result = predict(question="What is DSPy?")
print(result.answer)

Everything from DSPy works -- the import is still import dspy.

Loading images, audio, and files

Resource validation never reads the filesystem or network implicitly. Load resources explicitly:

image = dspy.Image.from_path("./image.png")
image = dspy.Image.from_url("https://example.com/image.png")  # explicit network request
image_reference = dspy.Image("https://example.com/image.png")  # provider-fetched URL; no local request

audio = dspy.Audio.from_path("./audio.wav")
file = dspy.File.from_path("./document.pdf")

Applications must validate or allowlist untrusted URLs before calling from_url(); explicit downloads do not provide SSRF filtering.

Disk-cache safety

Disk caches use restricted pickle deserialization by default. Custom cached classes must be registered with dspy.configure_cache(safe_types=[...]). Setting restrict_pickle=False restores unrestricted pickle compatibility, emits a warning, and must only be used when the cache directory is fully trusted. Unsafe or incompatible legacy entries are treated as cache misses rather than retried with unrestricted deserialization.

What changed from DSPy

	DSPy	dspy-lite
LM routing	litellm (34 MB, 201k LOC)	litelm (87 KB, 2.6k LOC)
numpy	core dep (60 MB)	optional `[embeddings]`
Clean install size	144 MB, 67 packages	24 MB, 36 packages
API	unchanged	unchanged

Documentation

Full DSPy documentation: dspy.ai

Development disclosure

This project is maintained with AI-assisted coding. Through 2026-05-14, I used Claude Code with Claude Opus 4.6/4.7. As of 2026-05-14, I use pi with GPT-5.5. The human maintainer remains responsible for what is merged and released.

License

MIT (inherited from DSPy)

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

dspy_lite-3.2.1.post2.tar.gz (273.0 kB view details)

Uploaded Sep 11, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

dspy_lite-3.2.1.post2-py3-none-any.whl (336.2 kB view details)

Uploaded Sep 11, 2026 Python 3

File details

Details for the file dspy_lite-3.2.1.post2.tar.gz.

File metadata

Download URL: dspy_lite-3.2.1.post2.tar.gz
Upload date: Sep 11, 2026
Size: 273.0 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for dspy_lite-3.2.1.post2.tar.gz
Algorithm	Hash digest
SHA256	`3b8ccb11c2a3028eea5f41ae6536730fd9c7fd12ce7c38357fc4f10fd60600c3`
MD5	`17519b3b8d7275a13c000b1b0fdf958c`
BLAKE2b-256	`e9aa77aff4642f193f537579a8a750987ecda21a15424be41b1f96ab2d692270`

See more details on using hashes here.

Provenance

The following attestation bundles were made for dspy_lite-3.2.1.post2.tar.gz:

Publisher: publish.yml on kennethwolters/dspy-lite

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: dspy_lite-3.2.1.post2.tar.gz
- Subject digest: 3b8ccb11c2a3028eea5f41ae6536730fd9c7fd12ce7c38357fc4f10fd60600c3
- Sigstore transparency entry: 2794161641
- Sigstore integration time: Sep 11, 2026
Source repository:
- Permalink: kennethwolters/dspy-lite@e9aa08ae46c95bbdb71edbd3adb9560e88e5e96f
- Branch / Tag: refs/tags/v3.2.1.post2
- Owner: https://github.com/kennethwolters
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish.yml@e9aa08ae46c95bbdb71edbd3adb9560e88e5e96f
- Trigger Event: release