Speed Modes

InstaHarvest v2 provides 4 speed modes that control concurrency, delays, and rate limiting.

Overview

Mode	Concurrency	Delay	Rate/min	Burst	Use Case
🐢 safe	5	0.8-2.0s	30	3	Personal account, ban-proof
⚡ fast	15	0.2-0.8s	60	8	Moderate scraping
🚀 turbo	50	0.05-0.3s	120	20	Heavy scraping + proxies
♾️ unlimited	1000	0s	∞	1000	Anonymous bulk scraping

Usage

Speed modes are async-only

The mode parameter is available only on AsyncInstagram. Sync Instagram always uses safe mode internally.

import asyncio
from instaharvest_v2 import AsyncInstagram

async def main():
    # Safe — default, ban-proof
    async with AsyncInstagram.from_env(mode="safe") as ig:
        user = await ig.users.get_by_username("nike")

    # Fast — 3x faster
    async with AsyncInstagram.from_env(mode="fast") as ig:
        user = await ig.users.get_by_username("nike")

    # Turbo — 10x faster with proxies
    async with AsyncInstagram.from_env(mode="turbo") as ig:
        ig.add_proxies(["http://p1:8080", "http://p2:8080"])
        user = await ig.users.get_by_username("nike")

    # Unlimited — anonymous, no limits
    async with AsyncInstagram.anonymous(unlimited=True) as ig:
        profile = await ig.public.get_profile("nike")

asyncio.run(main())

SpeedMode Parameters

from instaharvest_v2.speed_modes import SpeedMode

# Each mode is a frozen dataclass:
SpeedMode(
    name="safe",
    max_concurrency=5,          # Max parallel requests
    delay_range=(0.8, 2.0),     # Random delay between requests
    rate_per_minute=30,         # Token bucket rate limit
    burst_size=3,               # Fast requests before delay
    proxy_multiplier=3.0,       # +N concurrency per proxy
    error_backoff=2.0,          # Delay multiplier on errors
)

Proxy Scaling

When proxies are added, concurrency scales automatically:

Effective concurrency = base + (proxy_count × multiplier)

Mode	No proxy	2 proxies	5 proxies	10 proxies
safe	5	11	20	35
fast	15	25	40	65
turbo	50	70	100	150
unlimited	1000	1000	1000	1000

Hard cap

Proxy-based scaling has a hard cap of 200 concurrent requests (except unlimited).

Error Escalation

Speed modes include automatic error handling:

• Rate limit (429) → Escalation level +2
• Challenge/Checkpoint → Escalation level +3
• Other errors → Escalation level +1
• Success (30s streak) → Escalation level -1

Each escalation level increases delays by 30%.

Custom Mode

from instaharvest_v2.speed_modes import SpeedMode

custom = SpeedMode(
    name="custom",
    max_concurrency=25,
    delay_range=(0.3, 1.0),
    rate_per_minute=45,
    burst_size=5,
    proxy_multiplier=4.0,
    error_backoff=1.8,
)