Spanvero How it works Find a model Best by use Compare models Pricing

AI price trends — the independent record

Name: Spanvero AI price history
Creator: Spanvero
License: https://creativecommons.org/licenses/by/4.0/

Spanvero logs the full AI cost landscape once a day and never overwrites it — an append-only, version-controlled time-series of what AI really costs to run. We sell no compute, so we've no reason to bend the number. Here's the record so far.

Market pulse: over the last 56 days, the cheapest paid way to run AI held roughly flat. 33 price drops and 27 rises logged.

Monthly digest: State of Inference Costs — July 2026 →

Models tracked

406

GPU SKUs

125

Daily snapshots

Tracking since

June 20, 2026

Today's floor (August 14, 2026)

Cheapest paid API model: $0.02/1M blended — inclusionAI: Ling-2.6-flash
Cheapest GPU to rent: $0.06/hr — NVIDIA RTX 3060 12GB
Free to call (rate-limited): 18 models

The market floor, day by day

Date	Cheapest API	Cheapest GPU	Free	Models
August 14, 2026	$0.02/1M	$0.06/hr	18	406
August 13, 2026	$0.02/1M	$0.06/hr	18	404
August 12, 2026	$0.02/1M	$0.06/hr	19	401
August 11, 2026	$0.02/1M	$0.06/hr	17	397
August 10, 2026	$0.02/1M	$0.06/hr	17	395
August 9, 2026	$0.02/1M	$0.06/hr	17	395
August 8, 2026	$0.02/1M	$0.06/hr	17	395
August 7, 2026	$0.02/1M	$0.06/hr	17	395
August 6, 2026	$0.02/1M	$0.06/hr	17	335
August 5, 2026	$0.02/1M	$0.06/hr	17	333
August 4, 2026	$0.02/1M	$0.06/hr	17	333
August 3, 2026	$0.02/1M	$0.06/hr	17	332
August 2, 2026	$0.02/1M	$0.06/hr	17	332
August 1, 2026	$0.02/1M	$0.06/hr	17	331

Recent price moves

299 real day-over-day moves detected across the tracked window.

August 14, 2026: Google Gemini Flash Latest fell 75% to $1.125/1M (from $4.5)
August 14, 2026: DeepSeek: DeepSeek V4 Flash 0731 rose 62% to $0.21/1M (from $0.13)
August 14, 2026: Google: Gemini 3.6 Flash fell 50% to $2.25/1M (from $4.5)
August 14, 2026: Google: Gemini 3.6 Flash (batch) fell 50% to $1.125/1M (from $2.25)
August 14, 2026: Z.ai: GLM 5.2 fell 28% to $1.305/1M (from $1.825)
August 14, 2026: Qwen: Qwen3 VL 30B A3B Instruct fell 13% to $0.325/1M (from $0.375)
August 14, 2026: Qwen: Qwen3 Next 80B A3B Instruct rose 1% to $0.6/1M (from $0.595)
August 13, 2026: DeepSeek: DeepSeek V4 Pro rose 85% to $1.752/1M (from $0.9473)
August 13, 2026: Qwen: Qwen3.5-35B-A3B rose 32% to $0.75/1M (from $0.57)
August 13, 2026: Z.ai: GLM 4.6 fell 9% to $1.25/1M (from $1.375)
August 13, 2026: MoonshotAI: Kimi K2.7 Code fell 3% to $2.035/1M (from $2.1)
August 12, 2026: Qwen: Qwen3 VL 235B A22B Instruct fell 38% to $0.65/1M (from $1.055)
August 12, 2026: Z.ai: GLM 5.2 rose 15% to $1.825/1M (from $1.59)
August 12, 2026: DeepSeek: DeepSeek V3.1 Terminus fell 4% to $0.61/1M (from $0.635)
August 12, 2026: OpenAI: gpt-oss-120b fell 3% to $0.1/1M (from $0.1035)

The market index and price moves come from the live OpenRouter listing (the real, full market); per-model history is the dated price we publish for each model. We label which is which. How we stay honest →

Download / cite this data

The prices as portable files — same numbers, $0 markup, free to quote with a link back. Model prices as of 2026-08-10; GPU rates as of 2026-07-09.

prices.json · prices.csv · gpus.json · gpus.csv · the full feed →

Priced by Spanvero on 2026-08-10 — spanvero.com/data

Open the Outcome Lab → · All models · Compare

The weekly price index

A short email of real AI price moves, straight from the daily log — no hype. We're collecting the list now; the first issue goes out when it opens. Unsubscribe with one click.

Joining the list needs JavaScript — or just email support@spanvero.com and we'll add you.