# Compute Screener

> Free compute prices and price charts: daily price indices for renting GPUs (USD per GPU-hour) and for AI model tokens by company (USD per 1M tokens), GPU prices updated every minute, with charts, history and a published methodology.

GPU prices as of November 2, 2026 at 12:08 AM CT; token prices as of November 2, 2026 at 12:09 AM CT. Each link opens that GPU's or company's page with its chart, history, providers or models, and a description. The figures are licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/): please cite them as "Compute Screener (computescreener.com)" with the page's link.

## GPU rental prices (USD per GPU-hour; strongest 13-week price trend first, up or down)

- [L40](https://computescreener.com/gpu/l40): $0.97 per GPU-hour; 1D +0.8%, 1W +0.5%, 1M -0.5%; Strong uptrend, +3.3% a month (score +7.4)
- [H100](https://computescreener.com/gpu/h100): $3.82 per GPU-hour; 1D +3.1%, 1W +3.1%, 1M +4.4%; Strong uptrend, +2.9% a month (score +7.3)
- [B300](https://computescreener.com/gpu/b300): $9.04 per GPU-hour; 1D -1.1%, 1W -0.9%, 1M +6.6%; Strong uptrend, +6.7% a month (score +5.4)
- [H200](https://computescreener.com/gpu/h200): $5.04 per GPU-hour; 1D -0.1%, 1W -1.1%, 1M +9.5%; Strong uptrend, +4.7% a month (score +5.3)
- [RTX A6000](https://computescreener.com/gpu/rtx-a6000): $0.59 per GPU-hour; 1D +1.4%, 1W -0.5%, 1M +3.9%; Strong uptrend, +4.0% a month (score +4.7)
- [RTX 6000 Ada](https://computescreener.com/gpu/rtx-6000-ada): $0.92 per GPU-hour; 1D +2.1%, 1W -0.4%, 1M +8.1%; Strong uptrend, +8.4% a month (score +4.3)
- [B200](https://computescreener.com/gpu/b200): $8.18 per GPU-hour; 1D +3.5%, 1W +4.0%, 1M +23.7%; Moderate uptrend, +9.3% a month (score +3.8)
- [RTX PRO 6000](https://computescreener.com/gpu/rtx-pro-6000): $2.08 per GPU-hour; 1D +2.1%, 1W +3.9%, 1M +11.3%; Moderate uptrend, +2.4% a month (score +3.0)
- [A100](https://computescreener.com/gpu/a100): $1.76 per GPU-hour; 1D +0.3%, 1W +1.8%, 1M -9.1%; Moderate downtrend, -3.0% a month (score -2.6)
- [L40S](https://computescreener.com/gpu/l40s): $1.42 per GPU-hour; 1D +0.9%, 1W +0.4%, 1M +5.0%; Moderate uptrend, +3.0% a month (score +2.4)
- [RTX 4090](https://computescreener.com/gpu/rtx-4090): $0.54 per GPU-hour; 1D +0.5%, 1W -2.5%, 1M -5.0%; Moderate uptrend, +2.2% a month (score +2.3)

## AI token prices (USD per 1M tokens at the mix of agent tokens, 87% cache read, 12% input, 1% output; usage-weighted across each company's models)

- [Anthropic](https://computescreener.com/tokens/anthropic): $0.570 per 1M tokens; 1D +4.6%, 1W -6.4%, 1M -30.1%
- [DeepSeek](https://computescreener.com/tokens/deepseek): $0.0278 per 1M tokens; 1D -0.4%, 1W -0.7%, 1M -6.4%
- [StepFun](https://computescreener.com/tokens/stepfun): $0.218 per 1M tokens; 1D -0.8%, 1W -3.2%, 1M +210.5%
- [Alibaba](https://computescreener.com/tokens/alibaba): $0.0898 per 1M tokens; 1D +33.0%, 1W +8.3%, 1M +31.5%
- [Xiaomi](https://computescreener.com/tokens/xiaomi): $0.0342 per 1M tokens; 1D +2.7%, 1W -6.3%, 1M +30.5%
- [Z.ai](https://computescreener.com/tokens/z-ai): $0.147 per 1M tokens; 1D +11.2%, 1W +2.2%, 1M -6.8%
- [Google](https://computescreener.com/tokens/google): $0.236 per 1M tokens; 1D +1.7%, 1W +16.0%, 1M +65.9%
- [MiniMax](https://computescreener.com/tokens/minimax): $0.0422 per 1M tokens; 1D 0.0%, 1W -31.0%, 1M -57.9%
- [Meta](https://computescreener.com/tokens/meta): $0.0157 per 1M tokens; 1D 0.0%, 1W 0.0%, 1M 0.0%
- [OpenAI](https://computescreener.com/tokens/openai): $0.130 per 1M tokens; 1D -17.6%, 1W +52.5%, 1M -79.0%
- [Moonshot AI](https://computescreener.com/tokens/moonshot-ai): $0.581 per 1M tokens; 1D -7.6%, 1W -7.3%, 1M -24.6%
- [xAI](https://computescreener.com/tokens/xai): $0.735 per 1M tokens; 1D 0.0%, 1W 0.0%, 1M 0.0%
- [NVIDIA](https://computescreener.com/tokens/nvidia): $0.200 per 1M tokens; 1D 0.0%, 1W 0.0%, 1M 0.0%

## AI model API prices (USD per 1M tokens: posted input, cached input and output prices, and the price at the index's mix)

- [Claude Opus 5.5](https://computescreener.com/models/claude-opus-5-5) (Anthropic): $4.00 input, $0.20 cached input, $20.00 output, $0.854 at the mix
- [Claude Haiku 5.5](https://computescreener.com/models/claude-haiku-5-5) (Anthropic): $0.10 input, $0.01 cached input, $0.50 output, $0.0257 at the mix
- [Claude Sonnet 5.5](https://computescreener.com/models/claude-sonnet-5-5) (Anthropic): $2.00 input, $0.10 cached input, $10.00 output, $0.427 at the mix
- [Claude Sonnet 4.5](https://computescreener.com/models/claude-sonnet-4-5) (Anthropic): $3.00 input, $0.30 cached input, $15.00 output, $0.771 at the mix
- [GPT-6 Luna](https://computescreener.com/models/gpt-6-luna) (OpenAI): $0.10 input, $0.01 cached input, $0.50 output, $0.0257 at the mix
- [GPT-5.6 Luna](https://computescreener.com/models/gpt-5-6-luna) (OpenAI): $0.20 input, $0.02 cached input, $1.20 output, $0.0534 at the mix
- [GPT-6.1 Sol](https://computescreener.com/models/gpt-6-1-sol) (OpenAI): $2.00 input, $0.10 cached input, $10.00 output, $0.427 at the mix
- [GPT-6.1 Terra](https://computescreener.com/models/gpt-6-1-terra) (OpenAI): $1.00 input, $0.05 cached input, $5.00 output, $0.213 at the mix
- [GPT-6 Luna Pro](https://computescreener.com/models/gpt-6-luna-pro) (OpenAI): $0.10 input, $0.01 cached input, $0.50 output, $0.0257 at the mix
- [Gemini 3.8 Flash](https://computescreener.com/models/gemini-3-8-flash) (Google): $0.75 input, $0.075 cached input, $3.75 output, $0.193 at the mix
- [Gemini 4 Argon](https://computescreener.com/models/gemini-4-argon) (Google): $2.00 input, $0.10 cached input, $10.00 output, $0.427 at the mix
- [Gemini 3 Flash Preview](https://computescreener.com/models/gemini-3-flash-preview) (Google): $0.50 input, $0.05 cached input, $3.00 output, $0.134 at the mix
- [Gemini 3.1 Flash Lite](https://computescreener.com/models/gemini-3-1-flash-lite) (Google): $0.25 input, $0.025 cached input, $1.50 output, $0.0668 at the mix
- [Gemini 3.7 Flash](https://computescreener.com/models/gemini-3-7-flash) (Google): $0.75 input, $0.075 cached input, $3.75 output, $0.193 at the mix
- [Grok 4.7](https://computescreener.com/models/grok-4-7) (xAI): $2.00 input, $0.50 cached input, $6.00 output, $0.735 at the mix
- [Muse Spark 1.3 Contributor](https://computescreener.com/models/muse-spark-1-3-contributor) (Meta): $0.10 input, $0.002 cached input, $0.20 output, $0.0157 at the mix
- [Kimi K3.1 Pro](https://computescreener.com/models/kimi-k3-1-pro) (Moonshot AI): $2.00 input, $0.20 cached input, $10.00 output, $0.514 at the mix
- [Kimi K3](https://computescreener.com/models/kimi-k3) (Moonshot AI): $3.00 input, $0.30 cached input, $15.00 output, $0.771 at the mix
- [Qwen3.8 27B](https://computescreener.com/models/qwen3-8-27b) (Alibaba): $0.425 input, $0.085 cached input, $2.55 output, $0.150 at the mix
- [Qwen3.8 Flash](https://computescreener.com/models/qwen3-8-flash) (Alibaba): $0.15 input, $0.016 cached input, $0.47 output, $0.0366 at the mix
- [Qwen3.7 Flash](https://computescreener.com/models/qwen3-7-flash) (Alibaba): $0.03 input, $0.006 cached input, $0.13 output, $0.0101 at the mix
- [DeepSeek V4.1 Flash 1017](https://computescreener.com/models/deepseek-v4-1-flash-1017) (DeepSeek): $0.15 input, $0.003 cached input, $0.60 output, $0.0266 at the mix
- [DeepSeek V4 Pro 0813](https://computescreener.com/models/deepseek-v4-pro-0813) (DeepSeek): $0.66 input, $0.022 cached input, $1.98 output, $0.118 at the mix
- [GLM 5.3 Flash](https://computescreener.com/models/glm-5-3-flash) (Z.ai): $0.15 input, $0.03 cached input, $0.50 output, $0.0491 at the mix
- [GLM 5.3](https://computescreener.com/models/glm-5-3) (Z.ai): $1.40 input, $0.26 cached input, $4.40 output, $0.438 at the mix
- [GLM 5.2](https://computescreener.com/models/glm-5-2) (Z.ai): $1.40 input, $0.26 cached input, $4.40 output, $0.438 at the mix
- [MiMo-V2.6-Flash](https://computescreener.com/models/mimo-v2-6-flash) (Xiaomi): $0.14 input, $0.0028 cached input, $0.28 output, $0.0220 at the mix
- [MiMo-V2.6-Pro](https://computescreener.com/models/mimo-v2-6-pro) (Xiaomi): $0.435 input, $0.0036 cached input, $0.87 output, $0.0640 at the mix
- [Step 5 Preview](https://computescreener.com/models/step-5-preview) (StepFun): $1.00 input, $0.10 cached input, $2.70 output, $0.234 at the mix
- [Step 3.7 Flash](https://computescreener.com/models/step-3-7-flash) (StepFun): $0.20 input, $0.04 cached input, $1.15 output, $0.0703 at the mix
- [MiniMax M3.1 Flash](https://computescreener.com/models/minimax-m3-1-flash) (MiniMax): $0.15 input, $0.015 cached input, $0.60 output, $0.0370 at the mix
- [MiniMax M3](https://computescreener.com/models/minimax-m3) (MiniMax): $0.30 input, $0.06 cached input, $1.20 output, $0.100 at the mix
- [Mistral Large 4](https://computescreener.com/models/mistral-large-4-0) (Mistral): $0.70 input, $0.07 cached input, $2.10 output, $0.166 at the mix
- [Nemotron 3 Ultra](https://computescreener.com/models/nemotron-3-ultra-550b-a55b) (NVIDIA): $0.60 input, $0.12 cached input, $2.40 output, $0.200 at the mix
- [Claude Fable 5.1](https://computescreener.com/models/claude-fable-5-1) (Anthropic): $10.00 input, $0.25 cached input, $50.00 output, $1.92 at the mix
- [GPT-6 Astra](https://computescreener.com/models/gpt-6-astra) (OpenAI): $10.00 input, $1.00 cached input, $50.00 output, $2.57 at the mix
- [Muse Wonder](https://computescreener.com/models/muse-wonder) (Meta): $4.00 input, $0.20 cached input, $20.00 output, $0.854 at the mix
- [Qwen3.8 Max (0902)](https://computescreener.com/models/qwen3-8-max-0902) (Alibaba): $2.00 input, $0.25 cached input, $6.00 output, $0.517 at the mix
- [Muse Spark Open](https://computescreener.com/models/muse-spark-open) (Meta): $1.25 input, $0.15 cached input, $4.25 output, $0.323 at the mix
- [Beam 1.0 501B A23B](https://computescreener.com/models/beam-1-0-501b-a23b) (Reflection AI): $0.40 input, $0.04 cached input, $1.20 output, $0.0948 at the mix
- [Nemotron 3.5 Super VL](https://computescreener.com/models/nemotron-3-5-super-vl) (NVIDIA): $0.15 input, $0.02 cached input, $0.50 output, $0.0404 at the mix
- [Inkling](https://computescreener.com/models/inkling) (Thinking Machines): $1.00 input, $0.17 cached input, $4.05 output, $0.308 at the mix

## Compute futures prices (USD per GPU-hour: each exchange's own last price, and 24-hour volume)

- [H100 futures](https://computescreener.com/futures#h100): BingX H100 perpetual $2.917 (24h volume $356,894); OKX H100 perpetual $3.084 (24h volume $248,201); Bitget H100 perpetual $2.917 (24h volume $163,654); HTX H100 perpetual $2.917 (24h volume $25,529); Gate H100 perpetual $2.946 (24h volume $6,400); Lighter H100 perpetual $2.889 (24h volume $12); CME Group H100 rental futures (Launching: Awaiting CFTC approval); ICE H100 rental futures (Announced: No listing date yet); Architect AX H100 perpetual (Listed: Qualified institutions and individuals)
- [B200 futures](https://computescreener.com/futures#b200): BingX B200 perpetual $5.592 (24h volume $354,788); Bitget B200 perpetual $5.594 (24h volume $23,993); Gate B200 perpetual $5.656 (24h volume $5); CME Group B200 rental futures (Launching: Awaiting CFTC approval); ICE B200 rental futures (Announced: No listing date yet); Architect AX B200 perpetual (Planned: On Architect's roadmap)
- [H200 futures](https://computescreener.com/futures#h200): Architect AX H200 perpetual (Planned: On Architect's roadmap)
- [B300 futures](https://computescreener.com/futures#b300): Architect AX B300 perpetual (Planned: On Architect's roadmap)

## GPU rental prices by provider (USD per GPU-hour: each provider's lowest posted on-demand price)

- [Akash](https://computescreener.com/providers/akash): H200 from $5.95, H100 from $2.58, RTX PRO 6000 from $1.44, A100 from $1.49
- [BoostRun via Shadeform](https://computescreener.com/providers/boostrun): RTX 4090 from $0.49
- [Civo](https://computescreener.com/providers/civo): 
- [CoreWeave](https://computescreener.com/providers/coreweave): 
- [Crusoe via Shadeform](https://computescreener.com/providers/crusoe): A100 from $1.45, L40S from $1.50
- [Denvr via Shadeform](https://computescreener.com/providers/denvr): H100 from $2.50, A100 from $1.40
- [DigitalOcean](https://computescreener.com/providers/digitalocean): H100 from $4.41, A100 from $3.11
- [Excess Supply via Shadeform](https://computescreener.com/providers/excess-supply): B300 from $8.50, H200 from $4.00, A100 from $1.35, RTX 4090 from $0.40
- [Hyperstack](https://computescreener.com/providers/hyperstack): 
- [Jarvis Labs](https://computescreener.com/providers/jarvis-labs): 
- [Lambda](https://computescreener.com/providers/lambda): 
- [Latitude.sh](https://computescreener.com/providers/latitude-sh): H100 from $1.66
- [Lium](https://computescreener.com/providers/lium): B200 from $5.90, H200 from $2.99, RTX PRO 6000 from $1.60, RTX 6000 Ada from $0.80, RTX A6000 from $0.42, RTX 4090 from $0.35
- [Massed Compute](https://computescreener.com/providers/massed-compute): H100 from $2.73, A100 from $1.35, L40S from $0.88, L40 from $0.88, RTX 6000 Ada from $0.97, RTX A6000 from $0.57
- [Nebius](https://computescreener.com/providers/nebius): 
- [OVHcloud](https://computescreener.com/providers/ovhcloud): 
- [RunPod](https://computescreener.com/providers/runpod): B300 from $6.94, B200 from $5.98, H200 from $3.59, H100 from $1.99, RTX PRO 6000 from $1.64, A100 from $1.00, L40S from $0.79, L40 from $0.69, RTX 6000 Ada from $0.74, RTX A6000 from $0.33, RTX 4090 from $0.34
- [Scaleway](https://computescreener.com/providers/scaleway): B300 from $8.35, H100 from $3.20
- [Together AI](https://computescreener.com/providers/together-ai): 
- [Vast.ai](https://computescreener.com/providers/vast-ai): B200 from $7.50, H200 from $3.60, H100 from $2.54, RTX PRO 6000 from $1.07, A100 from $0.67, L40S from $0.73, L40 from $0.46, RTX 6000 Ada from $0.49, RTX A6000 from $0.45, RTX 4090 from $0.36
- [Verda (DataCrunch)](https://computescreener.com/providers/verda-datacrunch): B300 from $11.35, B200 from $8.87, H200 from $6.25, H100 from $4.79, RTX PRO 6000 from $2.65, A100 from $1.61, L40S from $1.90, RTX 6000 Ada from $1.50, RTX A6000 from $0.83

## More

- [AI token prices by company](https://computescreener.com/tokens): the token price index of every listed AI company, with changes and charts
- [GPU cloud prices by provider](https://computescreener.com/providers): each provider's lowest posted price for each GPU
- [Compute futures prices](https://computescreener.com/futures): every exchange's futures contracts on GPU rental prices, live
- [Every GPU's rental price](https://computescreener.com/gpus): RTX 5090, RTX 3090, RTX 3060, RTX 5060 Ti and 77 more GPUs, as posted today
- [Every AI model's API price](https://computescreener.com/models): 321 AI models from 52 companies, per 1M tokens as posted today
- [Wholesale electricity prices](https://computescreener.com/power): day-ahead prices at ERCOT's and the California ISO's trading hubs, the daily average of the hourly prices in USD per MWh, from 2017 in Texas and 2023 in California (ERCOT North $24.83 on 2026-11-02, CAISO SP15 $42.46 on 2026-11-02); data from ERCOT and the California ISO
- [AI coding tools by usage](https://computescreener.com/coding): the coding apps that send the most tokens through OpenRouter: the latest day, 7 and 30 days, and the change, with weekly tokens since January 2025 (Source: OpenRouter, openrouter.ai/rankings, CC BY 4.0)
- [AI data center sites](https://computescreener.com/datacenters): 14 AI data center campuses of GPU clouds and AI labs: capacity, users and status, each with its sources (the companies' releases and filings)
- [Methodology](https://computescreener.com/methodology): how each index is calculated, sources and rules
- [Free MCP, API and CSV](https://computescreener.com/api): every price as JSON (no key, CC BY 4.0) at https://computescreener.com/data/index.json and the other files listed there, daily CSV files, and the MCP connector
- [Get the app](https://computescreener.com/app): install Compute Screener on a phone's home screen or a computer, from the browser
- [Connect to AI apps](https://computescreener.com/connect): MCP server at https://computescreener.com/mcp (read-only, no sign-in) for Claude, ChatGPT, Grok and others