Unverified50% confidenceFactExact time
Vercel released its October AI Gateway production index: as of September, open-weight models on the platform ran 71.5% of tokens but accounted for only 23% of spending; DeepSeek alone accounted for 58% of tokens while taking only 6.6% of spending; V4.1 Flash alone handled nearly half the workload while accounting for only 2% of spending.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/18/2026
First Seen
Valid until: 11/1/2026
Sources
23 AI Technologies Slated for Export Control: How Are API and Open Source Defined?
bilibili绿鸭Lvya10/17/2026
Related Entities
Related Claims
UnverifiedDeepSeek-v4.1-Flash alone handled nearly half of September's gateway token volume, but accounted for only 2.6% of spend, with its price per token just one-fifth of the open-source average.74% similarUnverifiedDeepSeek's models handled 58% of all gateway tokens but accounted for only 6.6% of spend; Moonshot AI took 3.6% of tokens and 13.2% of spend; Z.ai took 7.2% of tokens and 4.5% of spend.69% similarUnverifiedDeepSeek V4.1 Flash L017 reaches a 77% pass rate on the DDSWE coding benchmark, at an average cost of only 1.3 yuan per task.67% similarUnverifiedRunning vLLM 0.27.1 official image to infer Qwen2.5-0.5B-Instruct (fp16) on an H100 SXM 80GB on vast.ai ($2.689 per hour, 20 vCPU), the cost per million output tokens is about $1.26 under a single-request serial load.67% similarUnverifiedOpen-weight models account for more than 71.5% of the token traffic handled by AI Gateway and about 23% of total spend; since June, their share of spend has more than quadrupled.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1059900API
curl https://kongchang.com/api/v1/knowledge/claims/1059900MCP
get_claim(id=1059900)