ABTestAPI

ig.ab_test — A/B test different captions, hashtags, and posting strategies to find what works best.

Quick Example

ig = Instagram.from_env()

# Create an A/B test
test = ig.ab_test.create("Caption Test", variants={
    "A": {"caption": "Short and sweet ✨"},
    "B": {"caption": "Detailed caption with storytelling and hashtags #test #content"},
})

# Record results after posts go live
ig.ab_test.record(test["id"], "A", likes=100, comments=20, saves=15)
ig.ab_test.record(test["id"], "B", likes=150, comments=30, saves=25)

# Get winner
result = ig.ab_test.results(test["id"])
print(f"Winner: Variant {result['winner']}")
print(f"Improvement: {result['improvement_pct']}%")

---

Methods

create(name, variants, metric="engagement")

Create a new A/B test.

Param	Type	Default	Description
name	str	—	Test name
variants	dict	—	Variant definitions
metric	str	"engagement"	"engagement", "likes", or "comments"

---

record(test_id, variant_name, media_id=None, likes=0, comments=0, reach=0, saves=0)

Record results for a variant.

---

collect(test_id)

Auto-collect live engagement data from Instagram for all variants with recorded media IDs.

ig.ab_test.collect(test["id"])  # Fetches current like/comment counts

---

results(test_id)

Analyze results and determine the winner.

Returns:

{
    "winner": "B",
    "scores": {"A": 120, "B": 180},
    "improvement_pct": 50.0,
    "metric": "engagement"
}

---

list_tests(status="")

List all tests, optionally filtered by status.

delete_test(test_id)

Delete a test permanently.

---

Async Version

async with AsyncInstagram.from_env() as ig:
    test = await ig.ab_test.create("Test", variants={"A": {}, "B": {}})
    await ig.ab_test.record(test["id"], "A", likes=100)
    result = await ig.ab_test.results(test["id"])