Skip to content
The Chat Archive
chatgpt-4o-latest-20250326· OpenAIHugging Face LMArena Human Preference 140K

what would be the cost

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with o4-mini-2025-04-16’s answer.
  1. User

    what would be the cost

  2. Assistant

    Great! Let’s walk through cost estimates for taking an open-source DeepSeek R1-style model and adapting it into a general-purpose foundational robot.


    🧮 Estimate Summary: Total Cost with DeepSeek R1

    PhaseEst. GPU Hours (H100)Cloud Cost (@$2/hr)On-Premise Cost (optional)
    Fine-tuning (domain/robot)1,000 – 20,0002,000–2,000 – 40,000700–700 – 14,000 (amortized)
    Sim2Real Transfer (training)10,000 – 100,00020,000–20,000 – 200,0007,000–7,000 – 70,000
    RL / Policy Optimization10,000 – 100,00020,000–20,000 – 200,0007,000–7,000 – 70,000
    Real-World Fine-Tuning / Updates1,000 – 10,0002,000–2,000 – 20,000700–700 – 7,000
    Total22,000 – 230,000 hrs44,000–44,000 – 460,00015,000–15,000 – 160,000

    ✅ Starting from DeepSeek R1 saves up to ~90–97% of the cost versus training from scratch (which would be 2M–2M–20M+).


    🔄 Cloud vs. On-Premise Costs

    💡 Cloud Pricing

    • Standard rate for H100 (80GB) on major clouds (Lambda, CoreWeave, AWS, etc.): 1.8–1.8–2.5/hr
    • For cost range, we use ~$2/hr baseline.

    ⚙️ On-Premise (Optional)

    If you set up your own training farm, amortized costs can be significantly lower:

    • Cost per H100: ~$25,000
    • Operating lifespan: 3 years (24/7 ≈ 26,000 hrs)
    • Net amortized cost/hr: ~0.80–0.80–1.20/hr (plus power, cooling, etc.)

    So if you own 4–8 H100s, and train efficiently, you could trim costs to ~15,000–15,000–160,000.


    🚗 Additional Deployment Costs (Optional)

    ComponentEstimated Cost
    Real-world robot hardware (e.g. mobile base, manipulators)20,000–20,000–200,000
    Sensors (LIDAR, RGB-D, IMU, force-torque):5,000–5,000–25,000
    Onboard compute (e.g., NVIDIA Orin, Thor, Jetson)2,000–2,000 – 30,000
    Infrastructure (teleop, data labeling, safety monitoring)10,000–10,000 – 100,000

    ✅ Cost Optimization Tips

    • Start in simulation: Extensively pre-train in sim (Isaac Gym, MuJoCo, RoboSuite) to avoid expensive real-world mistakes.
    • Use LoRA / QLoRA / DPO fine-tuning: Significantly reduces fine-tuning compute.
    • Quantize for inference (e.g., INT4): Shrinks model runtime costs by 2–4×.
    • Use MoE (Mixture of Experts): Activates only part of the model for a given input, saving compute.
    • Use distillation or splitting (e.g., multi-agent planners): Cheaper than end-to-end monolithic inference.

    👀 In Practice — Real-World Examples

    Company / ProjectEst. Training / Infra SpendDetails
    Tesla Optimus$500M+ (part of Dojo R&D)Heavy investment in data centers + real-world fleet
    Figure AI~$70M from OpenAI/MicrosoftBuilds on foundation models + industrial hardware
    DeepSeek RoboticsLikely <$5M (open weights)Pre-trained on large-scale vision/action/caption datasets
    Open-Source (💡you)44K–44K–460K (cloud)Starting from R1 + simulation + smart fine-tuning, you can prototype affordably

    📌 Bottom Line

    🚀 Total compute cost to build a general-purpose robot from DeepSeek R1:

    • Low-end (sim-heavy, few physical demos): ~50kcloud∗∗or∗∗ 50k cloud** or **~15k on-prem
    • Mid-range (real-world behaviors + adaptation): ~$200k
    • High-end (complex tasks, long RL, real-world deployment): ~$400k–460k

    💡 That's <5% of the cost of training a GPT-like robotics model from scratch.


    If you'd like, I can give you a custom budget plan based on your:

    • Robot hardware stack
    • Task complexity (e.g., navigation vs. manipulation)
    • Available GPUs or cloud credits
    • Timeline constraints

    Would that help?

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.