> ## Documentation Index > Fetch the complete documentation index at: https://docs.galadriel.co/llms.txt > Use this file to discover all available pages before exploring further. # Pricing Guide > Understanding marketplace pricing and cost optimization strategies # Pricing Guide Galadriel uses **marketplace-driven pricing** where supply and demand determine GPU costs through a transparent orderbook system. This guide explains how pricing works and strategies to save money. ## Core Pricing Model ### Market-Driven vs Fixed Pricing | Aspect | Fixed Pricing (Traditional Cloud) | **Marketplace (Galadriel)** | | ------------------ | --------------------------------- | --------------------------- | | Price discovery | Platform sets price | Market determines price ✓ | | Demand response | Fixed regardless of demand | Adjusts to demand ✓ | | Customer savings | No opportunity for deals | Can bid below market ✓ | | Price transparency | Opaque | Full orderbook visibility ✓ | **Result**: Lower average prices, better utilization, more supply. ## Platform Fee All transactions include a **10% platform fee**: ``` Example Transaction: ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Customer pays: 16 GPUs × 24h × $1.40/hr = $537.60 Platform fee (10%): +$53.76 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Total charge: $591.36 Provider receives (90%): $483.84 ``` **What the 10% covers:** * ✅ Orderbook matching engine * ✅ Payment processing (Stripe fees \~3%) * ✅ Kubernetes orchestration * ✅ Network setup (InfiniBand config) * ✅ Monitoring and support * ✅ Platform development and operations ## Typical Pricing Patterns ### By Time of Day ``` Typical H100 Pricing (us-west-1) ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Weekday Peak (9am-6pm): Supply: 500 GPUs Demand: 800 GPUs Price: $1.40-1.50/hr ⬆️ Weekday Off-Peak (6pm-9am): Supply: 500 GPUs Demand: 200 GPUs Price: $1.20-1.30/hr ⬇️ (10-15% cheaper) Weekends: Supply: 500 GPUs Demand: 300 GPUs Price: $1.25-1.35/hr ➡️ (5-10% cheaper) Conference Deadlines (e.g., NeurIPS): Supply: 500 GPUs Demand: 1500+ GPUs Price: $1.80-2.00/hr ⬆️⬆️ (40-60% surge) ``` ### Price History Example View historical prices: ```bash theme={null} galadriel prices --gpu-type h100 --zone us-west-1 --days 7 ``` Output: ``` H100 Price History (us-west-1) - Last 7 Days ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ $2.00 ┤ ╭╮ $1.90 ┤ │╰╮ $1.80 ┤ ╭╯ ╰╮ $1.70 ┤ ╭───╮ │ ╰╮ $1.60 ┤ ╭───╯ ╰───╮ ╭───╯ ╰─╮ $1.50 ┤ ╭───╯ ╰──────╯ ╰── $1.40 ┤ ────╯ $1.30 ┤ $1.20 ┤────── └┬────┬────┬────┬────┬────┬────┬────┬─── Mon Tue Wed Thu Fri Sat Sun Mon Current: $1.35/hr 7-day avg: $1.42/hr 7-day low: $1.20/hr (Sun 3am) 7-day high: $1.95/hr (Wed 2pm - Conference deadline) 💡 Tip: Best prices typically 11pm-6am (15-20% less demand) ``` ## Cost Optimization Strategies ### 1. Time-Shift Workloads (Save 30-40%) Run workloads during off-peak hours: ``` Instead of: Running during peak hours Do this: Schedule for off-peak (nights/weekends) Example: Peak price: $1.45/hr Off-peak price: $1.25/hr Savings: 14% Job: 16 GPUs × 12 hours Peak cost: $278.40 Off-peak cost: $240.00 Saved: $38.40 ``` **How to do it:** ```bash theme={null} # Flexible order - execute anytime in next 48h galadriel order buy \ --gpu-type h100 \ --gpus 16 \ --duration 12h \ --flex 48h \ --price 1.30 ``` ### 2. Use Limit Orders (Save 10-25%) Set your maximum price and wait: ``` Instead of: Market orders (instant but expensive) Do this: Limit orders below current market Example: Current market: $1.40/hr Your limit: $1.25/hr (11% below) Wait time: 2-4 hours Job: 8 GPUs × 24 hours Market cost: $268.80 Limit cost: $240.00 Saved: $28.80 Worth waiting? YES if not urgent ``` **How to do it:** ```bash theme={null} galadriel order buy \ --gpu-type h100 \ --gpus 8 \ --duration 24h \ --price 1.25 ``` ### 3. Monitor the Orderbook Check prices before buying: ```bash theme={null} # View current orderbook galadriel orderbook --gpu-type h100 --zone us-west-1 # Get price estimate galadriel prices estimate --gpu-type h100 --gpus 16 --zone us-west-1 ``` If the spread is wide or prices are high, consider: * Placing a limit order below best ask * Waiting for off-peak hours * Using flexible timing ### 4. Resell Unused Time Finished early? Recover costs by reselling: ``` Original purchase: 168h × 16 GPUs × $1.30 = $3,494.40 Used: 120h (5 days) Remaining: 48h (2 days) Without reselling: - Refund: 48h × 16 × $1.30 = $998.40 With reselling at $1.20/GPU/hr: - Refund: $998.40 (from Galadriel) - Resell revenue: 48h × 16 × $1.20 = $921.60 - Platform fee (10%): -$92.16 - Net recovery: $829.44 - Total back: $998.40 + $829.44 = $1,827.84 - Extra recovered: $829.44 ``` **How to do it:** ```bash theme={null} galadriel lease resell lse_xyz123 --price 1.20 ``` See the [Reselling Guide](/marketplace/reselling) for details. ## Comparing Order Types ### Cost Comparison Example Job: 16 GPUs × 24 hours in us-west-1 | Order Type | Price | Total Cost | Wait Time | Savings | | ------------------ | --------- | ---------- | --------- | ------------- | | **Market** | \$1.40/hr | \$537.60 | 0 min | Baseline | | **Limit (\$1.30)** | \$1.30/hr | \$499.20 | \~1-2 hrs | \$38.40 (7%) | | **Limit (\$1.25)** | \$1.25/hr | \$480.00 | \~3-5 hrs | \$57.60 (11%) | | **Flexible (48h)** | \$1.20/hr | \$460.80 | Scheduled | \$76.80 (14%) | Platform fee (10%) is included in all costs shown above. ### When to Use Each Order Type **Market Orders:** * ✅ Deadline-driven workloads * ✅ Production inference serving * ✅ Time-sensitive experiments * ❌ Cost-optimized training * ❌ Batch processing jobs **Limit Orders:** * ✅ Non-urgent training runs * ✅ Cost-sensitive workloads * ✅ Predictable schedules * ❌ Immediate needs * ❌ Guaranteed execution time **Flexible Orders:** * ✅ Maximum cost savings * ✅ Batch jobs * ✅ Dataset preprocessing * ✅ Model evaluation * ❌ Interactive development * ❌ Fixed deadlines ## Billing & Invoicing ### How Billing Works **1. Order Placement** ``` Customer creates buy order: 16 GPUs @ max $1.35/hr × 24h Platform holds authorization: $1.35 × 16 × 24 × 1.10 = $570.24 (Includes 10% platform fee) ``` **2. Order Matching** ``` Matched with provider at $1.30/hr (below max!) Actual charge: $1.30 × 16 × 24 × 1.10 = $548.48 Saved: $21.76 vs max price ``` **3. Early Termination** ``` If terminated early (e.g., after 20 hours): Actual usage: $1.30 × 16 × 20 × 1.10 = $457.60 Refund: $548.48 - $457.60 = $90.88 ``` ### Monthly Invoice Example ``` Galadriel GPU Marketplace - November 2025 Customer: alice@example.com ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ DATE ORDER GPUs HOURS RATE COMPUTE FEE TOTAL ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 11/05 14:00 ord_abc123 16 24.0 $1.30 $499.20 $49.92 $549.12 11/08 09:00 ord_def456 8 12.5 $1.45 145.00 14.50 159.50 11/15 22:00 ord_ghi789 32 48.0 $1.20 1,843.20 184.32 2,027.52 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Subtotal (Compute): $2,487.40 Platform Fees (10%): $248.74 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Total: $2,736.14 Payment Method: Card ending in 1234 Due Date: 2025-12-08 💡 Savings this month: $412 by using limit orders and off-peak hours ``` View billing: ```bash theme={null} galadriel billing usage galadriel billing invoices ``` ## What's Included (No Extra Fees) All instances include: * ✅ **Compute**: Full node exclusivity (8 GPUs minimum) * ✅ **InfiniBand**: 400 Gbps RDMA, GPUDirect enabled * ✅ **Network Egress**: Unlimited data transfer out * ✅ **Storage**: At least 2TB of local NVMe storage per node * ✅ **Support**: Email/Slack support, documentation * ✅ **API Access**: Unlimited API calls * ✅ **SSH Access**: Direct SSH to all your nodes * ✅ **Kubernetes Access**: Full kubectl access **No hidden fees. Just the 10% platform fee on compute.** ## Refund & Cancellation Policy ### Early Termination by Customer ``` Scenario: Ordered 24h, terminated after 4.5h Original order: 16 GPUs × $1.30/hr × 24h = $499.20 + $49.92 fee = $549.12 Actual usage: 16 GPUs × $1.30/hr × 4.5h = $93.60 + $9.36 fee = $102.96 Refund: $549.12 - $102.96 = $446.16 (credited immediately) ``` ### Cancellation Before Start ``` over 24 hours before: Full refund (100%) 1-24 hours before: 50% refund (provider gets 50% of 1hr cost) under 1 hour before: No refund (provider gets 1hr minimum) ``` **Rationale**: Protects providers who blocked capacity ### Hardware Failure (Platform Responsibility) ``` Job running 10 hours, GPU fails, 1 hour downtime Full refund for downtime: 1hr × 16 GPUs × $1.30 × 1.10 = $22.88 Automatic credit to account Provider still gets paid (platform absorbs cost) ``` ## Price Guarantees ### For Customers **1. No Surprise Charges** * Limit orders guarantee max price * Market orders show price before confirmation * No hidden fees beyond 10% platform fee **2. Price Lock** * Once matched, price is fixed for entire duration * No surge pricing mid-job * Early termination gets prorated refund at matched price **3. Price History** * All historical prices publicly visible * Can analyze trends before ordering * Algorithm recommendations based on patterns ### For Providers **1. Minimum Price Guarantee** * Your listing price is the minimum you'll accept * Platform never matches below your min price * Can adjust prices anytime (for future orders) **2. Payment Guarantee** * Platform holds customer payment upfront * You get paid even if customer cancels * Automatic payout after job completion ## Advanced: Price Estimation Get accurate price estimates before ordering: ```bash theme={null} # Estimate cost for market order galadriel prices estimate --gpu-type h100 --gpus 16 --zone us-west-1 ``` Output: ``` Price Estimate for Market Order ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ GPU Type: H100 Zone: us-west-1 Quantity: 16 GPUs Estimated price: $1.38/hr (weighted average) Best ask: $1.40/hr Worst fill price: $1.45/hr Total available: 48 GPUs Sufficient liquidity: YES ✓ For 24h lease: Estimated cost: $530.88 Platform fee: $53.09 Total: $583.97 ``` ## Next Steps Understand orderbooks and order types GPU types, system specs, and access model Recover costs by reselling unused time Real-world cost optimization examples