Guides → Playground & Guide → Robot Training-Data / Teleoperation Cost - Price the Demonstration Corpus

Robot Training-Data / Teleoperation Cost - Price the Demonstration Corpus

Meet Priya Anand. Robotics ML lead scoping a VLA fine-tune. "GPUs are cheap and available - but what does it actually cost to collect the demonstrations my policy needs, and can synthetic data cut that bill?"

🔥 Data is the bottleneck for embodied AI, not compute; a VLA fine-tune needs 300-1,200 demos/task at $118-200/hr fully-loaded, and quality-filtering throws away 20-30% - none of it visible in any calculator.

The story

Compute scaled; data did not. GPU clusters for billion-parameter VLAs are commercially available today - but the paired observation-action demonstrations those models need do not exist at internet scale. Human teleoperation is the highest-quality source, and it is expensive and slow.

This calculator prices that corpus. It sizes the demonstrations (tasks x demos/task), over-collects for quality-filter attrition, then adds collection labor, annotation, post-processing, rig amortization, and storage - reading the storage rate from the same resource-pricing SSOT behind aicost.ai's engines.

The sim-to-real lever is the story of 2026. Teams at CMU and Stanford independently found VLAs trained on 40% synthetic data matched 100% real on held-out tasks. Dial synthetic augmentation up to 40% and watch real collection - the dominant cost - fall.

Headline is the total data-program cost; the second hero is cost per usable demonstration. Two levers dominate: demos/task and the synthetic mix.

🎮 Playground

Robot Training-Data / Teleoperation Cost Playground

Here are the inputs that move the result the most. Play with the sliders and check it out. The number updates live.

What does the training-data program cost?

Drag demos/task, teleop rate, and synthetic augmentation to see the total data-program cost. The full calculator adds tasks, discard, annotation, rig, and storage.

Data-program cost

💡Demos/task sizes the corpus; teleop rate and throughput set collection labor; synthetic augmentation offsets real demos. The headline is the total program cost.

Ready to run the numbers?

Open the full calculator. Pick a model, enter your tokens, see per-call, daily, monthly, and annual cost.

🚀 Open the full calculator →

Top 3 right now

Verified 11 hours ago

About this calculator: Robot Training-Data / Teleoperation Cost - Price the Demonstration Corpus

Price the demonstration corpus to train a robot policy: teleoperation collection, quality discard, annotation, post-processing, rig, and storage - with a sim-to-real synthetic lever. Total program cost and cost per usable demo.

🎛 Inputs you control

Each input shapes the cost. Click an input on the calculator to set it. The explanations below match the live calculator field by field.

Tasks in program: Distinct manipulation tasks/skills to train.
How to choose: One for a single-skill pilot; several for a multi-task or generalist corpus.
Demos / task: Demonstration episodes per task.
How to choose: ACT/Diffusion single-task 50-200; VLA fine-tune (OpenVLA, pi0) 300-1,200 by complexity.
Teleop $/hr (fully-loaded): Fully-loaded teleoperation cost per hour (operator + hardware depreciation + compute + QA).
How to choose: 2026: $118-200/hr for high-quality; $15-30/hr for simple 2D programs.
Demos / hour (throughput): Collection throughput in episodes per hour.
How to choose: ~40/hr typical; UMI GoPro-gripper ~3x faster; dexterous bimanual slower.
Quality discard (%): Percent of recorded episodes discarded by quality filtering.
How to choose: 20-30% is typical; you must over-collect to hit your usable count.
Synthetic augmentation (%): Percent of real demos offset by simulated variants.
How to choose: Up to ~40% - 2026 results show 40% synthetic matched 100% real on held-out tasks.
Annotation $/episode: Language/task-instruction annotation per episode.
How to choose: $0.25-1.00 human; $0.05-0.10 VLM-assisted.
Post-processing $/episode: Segmentation, success labeling, format conversion (RLDS/HDF5) per episode.
How to choose: $0.50-1.00/episode of ongoing engineering time.
Teleop rig capital $: Capital cost of the teleoperation rig.
How to choose: ALOHA 2 <$20K, GELLO <$300, OpenArm DIY ~$8K, humanoid setups higher.
Rig lifetime (episodes): Episodes to amortize the rig capital over.
How to choose: A rig is reused; spread its cost over its lifetime episode count.
Storage months: Months to retain the dataset.
How to choose: Storage is cheap; egress on repeated training transfers adds up.
Dataset transfers (egress): Times the dataset is transferred to GPU instances during development.
How to choose: Each full transfer incurs egress at ~$0.09/GB.
Most popular

Everything above is the 80% case. The last 20% is where the money is.

The gaps we just listed are real, and they are the expensive ones: your actual prompts, your switching cost, your MLOps overhead. An AICost expert spends the hour on your AI and cloud costs, not a generic playbook. You leave with a written report: the way forward, in 30 days of concrete steps.

  • An hour with the people who built the engines
  • Report the same day
  • The fee credits toward any AICost plan
  • Two slots a week
Book a Solution Session: $299 → or $99 for small business →

Not sure yet? The $39 AICost Blueprint credits toward a Session, and the Session fee credits toward any plan. You never pay twice for the same ground. See all pricing →

Ready to run your own numbers?

You have seen the shape of it. Open the calculator with your model, your tokens, your volume.

🚀 Open the full calculator →

Methodology

Editorial gate
8-layer defense, see aicost.ai/ai-cost-economics
Last verified
7/27/2026, 8:00:00 PM

Author: Subu Vdaygiri, Founder & CEO of CloudIntelligence.ai. 17 years Fortune 100 (Ingram Micro, Siemens). Wharton CTO program · Kellogg CPO program · 10× AWS+Azure certified.

📖 Data sources & methodology 163 text models · 9 embeddings · 37 vision · 55 audio · 8 vector DBs across 10 vendor pages · last verified 2026-07-28

Methodology

  • All prices are USD per 1 million tokens, current as of 2026-07-28.
  • Vendor-published values have no mark. Inferred/extrapolated values are marked with * and listed below.
  • Batch API discounts are 50% off standard rates across providers that offer Batch mode.
  • Prompt caching discounts vary by provider (typically 80-90% off cached input tokens).
  • Regional data-residency surcharges (Anthropic 1.1x, OpenAI 1.1x, Google regional tiers) are NOT included in base rates.
  • Long-context pricing tiers apply when input exceeds model threshold.
  • Embedding prices are input-only (no output tokens generated).

Primary sources

Last-verified date is the most recent successful daily snapshot (aicost_pricing_snapshots) or, when no snapshot exists yet, the latest successful crawler run (aicost_crawler_runs). 10 of 10 vendors are currently verified. Aggregator services (TokenCost, AI Pricing Guru, etc.) are not listed.

Anthropic
2026-07-28
https://www.anthropic.com/pricing
Daily snapshot since Sep 2023 · 631 days captured
Anthropic Docs
2026-07-28
https://platform.claude.com/docs/en/about-claude/pricing
Daily snapshot since Sep 2023 · 631 days captured
OpenAI
2026-07-28
https://openai.com/api/pricing/
Daily snapshot since Sep 2023 · 632 days captured
Google AI
2026-07-28
https://ai.google.dev/gemini-api/docs/pricing
Daily snapshot since Dec 2023 · 607 days captured
Google Vertex
2026-07-28
https://cloud.google.com/vertex-ai/generative-ai/pricing
Daily snapshot since Dec 2023 · 607 days captured
DeepSeek
2026-07-28
https://api-docs.deepseek.com/quick_start/pricing
Daily snapshot since May 2024 · 546 days captured
xAI
2026-07-28
https://x.ai/api
Daily snapshot since Nov 2024 · 464 days captured
Mistral
2026-07-28
https://mistral.ai/pricing
Daily snapshot since Dec 2023 · 605 days captured
Cohere
2026-07-28
https://cohere.com/pricing
Daily snapshot since Sep 2023 · 631 days captured

Inferred values (marked with * in calculator tables)

Derived from industry conventions, not directly published by the vendor. Typical conventions: cached input = 10% of base (90% off), Batch API = 50% of base (50% off).

Vendor / Model Field Why it’s inferred
Anthropic — Claude Sonnet 4.6 cachedInput Derived at 10% of input rate — Anthropic publishes 90% cache-hit discount on this tier.
Anthropic — Claude Sonnet 4.5 cachedInput Derived at 10% of input rate; same 90% cache-hit convention as Sonnet 4.6.
Anthropic — Claude Sonnet 4.5 batchInput Derived at 50% of standard input — Anthropic documents uniform 50% Batch discount.
Anthropic — Claude Sonnet 4.5 batchOutput Derived at 50% of standard output — Anthropic documents uniform 50% Batch discount.
Anthropic — Claude Haiku 4.5 cachedInput Derived at 10% of input rate — Anthropic 90% cache-hit discount convention.
OpenAI — GPT-5.4 Mini cachedInput Derived at 10% of input — OpenAI documents automatic 90% discount on cache hits across GPT-5.x tier.
OpenAI — GPT-5.4 Nano cachedInput Derived at 10% of input — OpenAI 90% cache-hit convention.
OpenAI — GPT-5.4 Nano batchInput Derived at 50% of input — OpenAI Batch API uniform 50% discount.
OpenAI — GPT-5.4 Nano batchOutput Derived at 50% of output — OpenAI Batch API uniform 50% discount.
OpenAI — GPT-5.4 Pro cachedInput Derived at 10% of input — OpenAI 90% cache-hit convention.
OpenAI — GPT-5.4 Pro batchInput Derived at 50% of input — OpenAI Batch API uniform 50% discount.
OpenAI — GPT-5.4 Pro batchOutput Derived at 50% of output — OpenAI Batch API uniform 50% discount.
OpenAI — GPT-5.2 cachedInput Derived at 10% of input; no residency uplift.
OpenAI — GPT-5.2 batchInput Derived at 50% of input.
OpenAI — GPT-5.2 batchOutput Derived at 50% of output.
OpenAI — GPT-5 cachedInput Derived at 10% of input.
OpenAI — GPT-5 batchInput Derived at 50% of input.
OpenAI — GPT-5 batchOutput Derived at 50% of output.
OpenAI — GPT-5.5 Pro cachedInput Derived at 10% of input — OpenAI does not publish a cached rate for *-pro models; using the family convention.
OpenAI — GPT-5.5 Pro batchInput Derived at 50% of input.
OpenAI — GPT-5.5 Pro batchOutput Derived at 50% of output.
OpenAI — GPT-5.2 Pro cachedInput Derived at 10% of input — pro-tier convention.
OpenAI — GPT-5.2 Pro batchInput Derived at 50% of input.
OpenAI — GPT-5.2 Pro batchOutput Derived at 50% of output.
OpenAI — GPT-5.1 batchInput Derived at 50% of input.
OpenAI — GPT-5.1 batchOutput Derived at 50% of output.
OpenAI — GPT-5 Pro batchInput Derived at 50% of input.
OpenAI — GPT-5 Pro batchOutput Derived at 50% of output.
OpenAI — GPT-5 Nano cachedInput Derived at 10% of input.
OpenAI — GPT-5 Nano batchInput Derived at 50% of input.
OpenAI — GPT-5 Nano batchOutput Derived at 50% of output.
Google — Gemini 3 Flash cachedInput Derived at 10% of input — Google caching discount convention ~90%.
Google — Gemini 3.1 Flash-Lite cachedInput Derived at 10% of input — Google caching convention.
Google — Gemini 3.1 Flash-Lite batchInput Derived at 50% of input — Google Batch API uniform 50% discount.
Google — Gemini 3.1 Flash-Lite batchOutput Derived at 50% of output — Google Batch API uniform 50% discount.
Google — Gemini 2.5 Pro cachedInput Derived at 10% of input.
Google — Gemini 2.5 Flash cachedInput Derived at 10% of input.
Google — Gemini 2.5 Flash-Lite cachedInput Derived at 10% of input — Google caching convention.
Google — Gemini 2.5 Flash-Lite batchInput Derived at 50% of input — Google Batch API uniform 50% discount.
Google — Gemini 2.5 Flash-Lite batchOutput Derived at 50% of output — Google Batch API uniform 50% discount.
Google — Gemini 2.0 Flash cachedInput Derived at 25% of input per Google 2.0 family caching rates.
Google — Gemini 2.0 Flash batchInput Derived at 50% of input — Google Batch API uniform 50% discount.
Google — Gemini 2.0 Flash batchOutput Derived at 50% of output — Google Batch API uniform 50% discount.
Google — Gemini 2.0 Flash-Lite cachedInput Derived at 10% of input — Google caching convention.
Google — Gemini 2.0 Flash-Lite batchInput Derived at 50% of input — Google Batch API uniform 50% discount.
Google — Gemini 2.0 Flash-Lite batchOutput Derived at 50% of output — Google Batch API uniform 50% discount.
xAI — Grok 4 (legacy) cachedInput Extrapolated at 25% of base.

Pricing is cross-verified against the LiteLLM community registry when available. Daily snapshots are kept in aicost_pricing_snapshots; every change is logged to aicost_price_changelog with old & new values for full audit trail. Read the full methodology →