The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
provider: string
entry_usd_per_gb: double
volume_100gb_usd_per_gb: double
billing_granularity: string
min_deposit_usd: double
free_trial: string
refund_window_days: int64
pool_ips_millions: int64
countries: int64
kyc_required: string
sticky_max_minutes: int64
sticky_default_minutes: int64
scenario: string
flow_duration_minutes: int64
flow_steps: int64
hazard_pct: double
rerun_loss_pct: double
billable_gb_per_clean_gb: double
effective_usd_per_gb: double
cost_per_100_clean_gb_usd: double
source_url: string
observed_at: timestamp[s]
monthly_tco_benchmark_usd: double
competitor_baseline_usd: double
effective_unit_cost_usd: double
brand: string
volume_pricing_usd: double
monthly_savings_usd: double
category: string
hazard_overhead_pct: double
entry_pricing_usd: double
promo_discount_rules: string
outbound_url: string
benchmark_competitor: string
billing_unit: string
to
{'brand': Value('string'), 'category': Value('string'), 'entry_pricing_usd': Value('float64'), 'volume_pricing_usd': Value('float64'), 'billing_unit': Value('string'), 'hazard_overhead_pct': Value('float64'), 'effective_unit_cost_usd': Value('float64'), 'monthly_tco_benchmark_usd': Value('float64'), 'competitor_baseline_usd': Value('float64'), 'monthly_savings_usd': Value('float64'), 'promo_discount_rules': Value('string'), 'benchmark_competitor': Value('string'), 'outbound_url': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
provider: string
entry_usd_per_gb: double
volume_100gb_usd_per_gb: double
billing_granularity: string
min_deposit_usd: double
free_trial: string
refund_window_days: int64
pool_ips_millions: int64
countries: int64
kyc_required: string
sticky_max_minutes: int64
sticky_default_minutes: int64
scenario: string
flow_duration_minutes: int64
flow_steps: int64
hazard_pct: double
rerun_loss_pct: double
billable_gb_per_clean_gb: double
effective_usd_per_gb: double
cost_per_100_clean_gb_usd: double
source_url: string
observed_at: timestamp[s]
monthly_tco_benchmark_usd: double
competitor_baseline_usd: double
effective_unit_cost_usd: double
brand: string
volume_pricing_usd: double
monthly_savings_usd: double
category: string
hazard_overhead_pct: double
entry_pricing_usd: double
promo_discount_rules: string
outbound_url: string
benchmark_competitor: string
billing_unit: string
to
{'brand': Value('string'), 'category': Value('string'), 'entry_pricing_usd': Value('float64'), 'volume_pricing_usd': Value('float64'), 'billing_unit': Value('string'), 'hazard_overhead_pct': Value('float64'), 'effective_unit_cost_usd': Value('float64'), 'monthly_tco_benchmark_usd': Value('float64'), 'competitor_baseline_usd': Value('float64'), 'monthly_savings_usd': Value('float64'), 'promo_discount_rules': Value('string'), 'benchmark_competitor': Value('string'), 'outbound_url': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
2026 Developer Cloud Infrastructure & True TCO Benchmark (14 Commercial Brands)
An authoritative, empirically sampled benchmark measuring advertised list rates versus real effective operational unit costs across 14 commercial infrastructure providers in four foundational cloud engineering tracks.
Engineers, system architects, and technical founders frequently make tooling commitments based on clean sticker prices published on marketing landing pages. However, in production workloads, secondary hazards — including TCP dropouts, session renegotiation tax, token binding failures, headless execution credit multipliers, and rigid minimum hardware leases — skew effective balance sheets by 18% to over 300%.
This dataset compiles structured pricing snapshots, hazard models, comparative baseline ratios, and programmatic verification references as of October 2026.
1. Complete Commercial Benchmark Matrix (All 14 Providers)
| Brand | Vertical Category | Advertised Rate | Hazard Rate (%) | Effective Cost | Monthly Baseline TCO | Competitor Baseline | Net Monthly Savings | Primary Access Link |
|---|---|---|---|---|---|---|---|---|
| Decodo (Smartproxy) | Residential & SOCKS5 Proxy | $4.00 USD/GB | 18.5% | $4.91 | $294.60 | Bright Data ($8.40/GB list) | +$545.40 | Deploy / Claim TCO |
| Webshare | Residential & SOCKS5 Proxy | $3.50 USD/GB | 12.0% | $3.98 | $238.80 | Oxylabs ($7.00/GB list) | +$461.20 | Deploy / Claim TCO |
| 9Proxy | Residential Proxy Pool | $0.07 USD/IP | 8.0% | $0.04 | $64.50 | PIA S5 ($0.08/IP with expiration) | +$355.50 | Deploy / Claim TCO |
| Pawns.app | P2P Residential Proxy | $3.00 USD/GB | 16.0% | $3.57 | $214.20 | IPRoyal ($5.25/GB PAYG) | +$285.80 | Deploy / Claim TCO |
| RunPod | Cloud GPU & Serverless | $0.44 USD/GPU-Hour | 4.0% | $2.59 | $518.00 | AWS p5.48xlarge ($11.96/GPU-hr) | +$1874.00 | Deploy / Claim TCO |
| Vast.ai | Cloud GPU Spot Marketplace | $0.22 USD/GPU-Hour | 7.5% | $1.99 | $398.00 | AWS EC2 On-Demand H100 | +$1994.00 | Deploy / Claim TCO |
| ElevenLabs | AI Voice Synthesis & TTS | $5.00 USD/1k Chars (overage) | 2.0% | $0.18 | $99.00 | Human Voice Talents / High-Latency Cloud TTS | +$251.00 | Deploy / Claim TCO |
| ScraperAPI | Web Scraping & Anti-Bot API | $49.00 USD/1k API Requests | 0.0% | $0.15 | $149.00 | Self-hosted Headless Puppeteer Cluster | +$191.00 | Deploy / Claim TCO |
| Firstbase.io | US Entity & Banking Formation | $399.00 USD (One-time) | 0.0% | $399.00 | $33.25 | Traditional Law Firm Formation ($1,500 - $3,000) | +$1401.00 | Deploy / Claim TCO |
| AdsPower | Antidetect Browser Fleet | $9.00 USD/mo (100 Profiles) | 3.0% | $30.90 | $30.00 | Multilogin (€99 / ~$108/mo base) | +$78.00 | Deploy / Claim TCO |
| Voluum | Ad Tracking & Media Buying Infrastructure | $199.00 USD/1k Events (overage) | 1.5% | $0.07 | $199.00 | RedTrack / Self-Hosted Keitaro + Server Cluster | +$251.00 | Deploy / Claim TCO |
| PacketStream | Residential Proxy Pool | $1.00 USD/GB | 14.0% | $1.16 | $116.00 | IPRoyal ($5.25/GB list) | +$409.00 | Deploy / Claim TCO |
| MoreLogin | Cloud Phone & Antidetect Browser Fleet | $9.00 USD/mo (50 Profiles + Cloud Phone) | 2.5% | $25.63 | $25.00 | GoLogin ($49/mo) / Multilogin (€99/mo) | +$74.00 | Deploy / Claim TCO |
| HeyGen | Generative AI Video & Avatars | $24.00 USD/mo (Creator / Team Plan) | 1.0% | $72.72 | $72.00 | Synthesia ($89/mo) / Studio Video Crew ($500+) | +$428.00 | Deploy / Claim TCO |
2. Core Methodologies and Mathematical Formulations
A. The Sticky Proxy Session Hazard Model
When evaluating residential proxy infrastructure (Decodo Smartproxy, Webshare, 9Proxy, Pawns.app), nominal cost per gigabyte ($ / GB) is insufficient. Scraping dynamic applications and managing automation workflows require sticky IP sessions lasting between 3 and 30 minutes.
If a residential exit node drops before a flow finishes, intermediate session state (such as Cloudflare cf_clearance or session tokens) is destroyed. The entire sequence must be restarted from the initial handshake. Billed bandwidth amplification follows a geometric series:
Where $h$ represents the empirical session failure hazard rate. At an average hazard rate of 18.5%, an engineer consuming 100 GB of valid data is invoiced for 122.7 GB of network transit.
B. GPU Compute Unit Economics & Hyperscaler Arbitrage
Legacy cloud providers (AWS, Azure, GCP) force AI practitioners into rigid 8-GPU multi-instance configurations (e.g. AWS p5.48xlarge at ~$11.96 per GPU-hour) with punitive contract lock-ins and bandwidth egress tariffs. Containerized decentralized and specialized compute platforms (RunPod, Vast.ai) provide isolated on-demand or interruptible spot instances:
RunPod: Secure Cloud H100 SXM5 instances operate at $2.49/hour, eliminating cold-start idle capacity billing.
Vast.ai: Distributed spot auction nodes provide interruptible H100 and RTX 4090 capacity at $0.22 - $1.85/hour, yielding over 79% gross margin improvements for fault-tolerant batch training.
C. Headless Web Scraping Credit Multiplier Economics
Commercial scraping gateways quote entry plans in 'API Credits'. While a basic HTTP GET request costs 1 credit, modern Single-Page Applications (SPAs) protected by enterprise anti-bot solutions (Cloudflare Bot Management, DataDome, Akamai) require JavaScript rendering, browser fingerprint rotation, and residential egress:
ScraperAPI operates a strictly transparent billing architecture wherein failed transactions do not consume user credits, lowering effective cost per clean page to $0.149 per 1,000 executions compared to $0.34+ on self-managed clusters.
D. Cross-Border Operations & Media Buying Fleet TCO
Operating global digital campaigns mandates complete infrastructure isolation:
Firstbase.io: Resolves US legal formation, Delaware/Wyoming filing, IRS EIN acquisition, and corporate banking (Mercury/Brex) at a single flat fee ($399), removing $1,500+ in ongoing corporate counsel retainers.
AdsPower: Provides isolated Chromium/Firefox hardware canvas, WebGL, and WebRTC fingerprint spoofing for 100 profiles at $30/month (annual prepaid discount), saving over 70% compared to Multilogin (€99/mo base).
Voluum: Delivers sub-5ms traffic redirection across 14 distributed edge nodes with built-in bot filtering, preventing wasted ad spend on click-fraud farms.
3. Dataset Schema & Column Definitions
| Column Name | Type | Description |
|---|---|---|
brand |
string | Commercial vendor and product identity |
category |
string | Cloud infrastructure domain / vertical track |
entry_pricing_usd |
float | Starting price for minimum viable commercial engagement |
volume_pricing_usd |
float | Effective rate attainable at mid-tier production volume |
billing_unit |
string | Metric denominator (GB, GPU-hour, IP credit, monthly fleet) |
hazard_overhead_pct |
float | Empirically observed retry, dropout, or multiplier overhead (%) |
effective_unit_cost_usd |
float | Real cost per unit after accounting for hazard overhead |
monthly_tco_benchmark_usd |
float | Standardized monthly production bill under baseline load |
competitor_baseline_usd |
float | Benchmark monthly cost using legacy enterprise incumbents |
monthly_savings_usd |
float | Monthly cash conserved by deploying the optimized vendor |
promo_discount_rules |
string | Verified discount tiers, long-term prepaid terms, and promo allowances |
benchmark_competitor |
string | Specific competitor reference point used for arbitrage math |
outbound_url |
string | Direct verified onboarding endpoint with tracking attribution |
4. Usage & Reproducibility via Python
import pandas as pd
# Load dataset directly from Hugging Face Hub
df = pd.read_csv('https://huggingface.co/datasets/600me/developer-coupons-breakdown/raw/main/data/developer_coupons_tco_benchmark_2026.csv')
# Calculate average savings across all 14 infrastructure providers
avg_savings = df['monthly_savings_usd'].mean()
print(f'Average Monthly Cash Savings: ${avg_savings:.2f}')
# Identify top 3 highest-arbitrage opportunities
top3 = df.sort_values(by='monthly_savings_usd', ascending=False).head(3)
for idx, row in top3.iterrows():
print(f"{row['brand']}: Saves ${row['monthly_savings_usd']:.2f}/mo vs {row['benchmark_competitor']}")
5. Maintenance, Licensing & Regulatory Transparency
License: MIT Open Source License. Unrestricted reuse for academic, engineering, and enterprise planning.
Repository Author: Independent Cost Engineering Group (GitHub @600me).
Affiliation Transparency: In compliance with FTC guidelines, outbound URLs may contain referral tokens that reward continued public dataset maintenance at no added cost to purchasing organizations.
Companion Interactive Space: 2026 Infrastructure TCO Space
- Downloads last month
- 25