Statistics · updated daily

AI model statistics, August 9, 2026

Every statistic below is computed directly from the same real data behind the rest of this site - never hand-written - so it updates automatically and never drifts. Free to cite with attribution; link directly to any numbered stat with its anchor.

1. 01

   Claude Opus 5 (batch) leads on real quality benchmarks with an Artificial Analysis Intelligence Index of 63.1, among 85 models with published scores, as of August 9, 2026.

   #quality-leaderFull data →
2. 02

   DeepSeek: DeepSeek V4 Flash 0423 handles the largest real share of production traffic on OpenRouter, at 12.9% of tracked token volume, averaged over the trailing 35 days.

   #usage-leaderFull data →
3. 03

   The median Intelligence Index across 85 benchmarked models is 37.4 - half of tracked models score higher, half lower.

   #median-quality
4. 04

   inclusionAI: Ling-2.6-flash is the cheapest tracked model at $0.0150 per million blended tokens.

   #cheapestFull data →
5. 05

   OpenAI: o1-pro (batch) is the most expensive tracked model at $262.50 per million blended tokens - a 17500x spread from the cheapest.

   #most-expensiveFull data →
6. 06

   The median blended price across tracked models is $0.85 per million tokens.

   #median-price
7. 07

   Ling-3.0-flash delivers the best measured quality per dollar, at 1200.0 intelligence-index points per dollar.

   #best-valueFull data →
8. 08

   Open-weight models handle 74% of real token volume on OpenRouter, as of August 8, 2026.

   #open-shareFull data →
9. 09

   Anthropic models make up 6% of tracked models by model count.

   #anthropic-shareFull data →
10. 10

    OpenAI models make up 19% of tracked models by model count.

    #openai-shareFull data →
11. 11

    Real production usage is classified across 29distinct task categories, measured directly from OpenRouter's classified traffic.

    #task-categoriesFull data →
12. 12

    DeepSeek: DeepSeek V4 Flash 0423 leads classification with 15% of classified production usage, as of August 9, 2026.

    #classification_taggingFull data →
13. 13

    DeepSeek: DeepSeek V4 Flash 0423 leads data extraction with 14% of classified production usage, as of August 9, 2026.

    #data-extractionFull data →
14. 14

    DeepSeek: DeepSeek V4 Flash 0423 leads roleplay & fiction with 47% of classified production usage, as of August 9, 2026.

    #roleplay_fictionFull data →
15. 15

    DeepSeek: DeepSeek V4 Flash 0423 leads workflow execution with 21% of classified production usage, as of August 9, 2026.

    #agent-workflow_executionFull data →
16. 16

    DeepSeek: DeepSeek V4 Flash 0423 leads content writing with 22% of classified production usage, as of August 9, 2026.

    #content_writingFull data →
17. 17

    DeepSeek: DeepSeek V4 Flash 0423 leads data transformation with 22% of classified production usage, as of August 9, 2026.

    #data-transformationFull data →
18. 18

    DeepSeek: DeepSeek V4 Flash 0423 leads conversation with 22% of classified production usage, as of August 9, 2026.

    #conversational_replyFull data →
19. 19

    DeepSeek: DeepSeek V4 Flash 0423 leads q&a & knowledge with 20% of classified production usage, as of August 9, 2026.

    #qa_knowledgeFull data →

Real usage vs. real quality

Average usage share (trailing 35 days) plotted against Artificial Analysis quality score, for every model with both. No composite score, just two real numbers side by side - the gap between them is the whole story: Claude Opus 5 (batch) tops quality but ranks #16 by usage; DeepSeek: DeepSeek V4 Flash 0423 tops usage with a quality score of 51.8.