Token Demand Index

Tracked inference volume hit 70.2 trillion tokens per week as of 2026-08-12, pushing the Token Demand Index to 298.4, up 64 percent over 30 days and 198 percent since 2026-05-06. The demand-weighted effective price is $1.20 per 1M tokens, while open-source models handle 69 percent of tracked traffic. That mix shows the market expanding through cheap, commoditized capacity rather than margins getting crushed across the board. It is the output-side companion to the Compute Tightness Index (the GPU input).

297.8Demand index (100 = 2026-05-06)
70.0TTokens / week
▲ +64%Demand, 30 days
69.2%Open-source share
$1.19Effective $/1M
$83.6MImplied spend/week*

As of 2026-08-13, tracked AI inference runs at 70.0T tokens/week — up +64% in 30 days and +198% since 2026-05-06. Open-source carries 69.2% of those tokens, holding the demand-weighted price near $1.19/1M — a fraction of frontier list prices even as frontier models themselves got pricier.

⤓ Download full daily series (CSV)

Token demand vs the effective price of inference

Total tokens/weekEffective $/1M (right)

Who consumes the tokens

Open-sourceOtherClaudeOpenAIGooglexAI

By provider group — latest snapshot

Click a group to expand its top contributing models.

GroupBlended $/1MTokens/weekShare30d demand
Open-source$0.4548.4T69.2%▲ +71%
Other$0.952.1T3%▲ +241%
Claude$6.005.2T7.4%▼ -28%
OpenAI$2.618.9T12.7%▲ +197%
Google$0.994.9T7%▲ +45%
xAI$1.56515.9B0.7%▲ +34%

Open-source = aggregated weights-available labs (DeepSeek, Qwen, Llama, Mistral, etc.). "Other" = unclassified OpenRouter providers. Share = % of tracked tokens. *Implied spend = list price × tokens, not realized revenue.

Methodology

Built from OpenRouter's model rankings + pricing, grouped into Claude / OpenAI / Google / xAI / Open-source / Other. Token counts are OpenRouter's trailing-7-day figures (the rankings "week" view) — i.e. each point is tokens processed over the prior week, not a single day. Per group: blended median $/1M (75/25 input/output) and weekly total_tokens. Demand index = latest weekly total ÷ the first week's total × 100. Effective $/1M = demand-weighted blended price (as-of prices carried forward across snapshots where only one side updated). Series since 2026-05-06; it grows as new snapshots land. Full history as CSV. Coverage is OpenRouter traffic — a large but partial slice of the market.

Pair it with the Compute Tightness Index (GPU input) for both sides of the AI-compute economy. Query the raw data from an agent via the Bargo MCP endpoint (get_inference_economics).