AI API & GPU price report: July 2026
In July 2026 we recorded 53 price changes across AI model APIs and GPU clouds: 2 API price drops, 7 increases and 44 GPU rental price changes. The biggest cut was GLM 5.1 on OpenRouter (−1%).
Price changes
53
API drops
2
API increases
7
H100 lowest
$1.68/hr
GPU price index
On-demand price per GPU-hour across the GPU clouds we track, at the end of July 2026, compared with the start of the month.
| GPU | Lowest | Cheapest on | Median | Median vs start | Offers |
|---|---|---|---|---|---|
| B300 | $7.39 | RunPod | $7.45 | no change | 2 |
| B200 | $3.50 | Vultr | $6.11 | −6% | 7 |
| H200 | $3.29 | Vast.ai | $4.14 | +4% | 8 |
| H100 | $1.68 | Latitude.sh | $3.27 | +43% | 20 |
| A100 80GB | $1.35 | Hyperstack | $1.58 | +1.0% | 14 |
| L40S | $0.99 | RunPod | $1.50 | +50% | 5 |
| RTX PRO 6000 | $1.07 | Vast.ai | $1.89 | +5% | 7 |
| RTX 5090 | $0.43 | Vast.ai | $0.71 | +137% | 2 |
| RTX 4090 | $0.35 | Vast.ai | $0.51 | +46% | 4 |
| MI300X | $1.85 | Vultr | $2.65 | no change | 2 |
Biggest API price drops
Price drops(2)
- GLM 5.1onOpenRouterinput $0.98 → $0.966/1M, output $3.08 → $3.036/1M↓−1%
- GLM 5.2onOpenRouterinput $0.94 → $0.93/1M↓−1%
Biggest API price increases
Price increases(7)
- Qwen 3 8BonOpenRouterinput $0.05 → $0.117/1M, output $0.4 → $0.455/1M↑+134%
- Qwen3 VL 8B InstructonOpenRouterinput $0.08 → $0.117/1M, output $0.5 → $0.455/1M↑+46%
- MiniMax M2.1onOpenRouterinput $0.29 → $0.3/1M, output $0.95 → $1.2/1M↑+26%
- Kimi K2.6onOpenRouterinput $0.55 → $0.66/1M, output $3.2 → $3.41/1M↑+20%
- DeepSeek V3onOpenRouterinput $0.2 → $0.24/1M, output $0.77 → $0.9/1M↑+20%
- Qwen3.6 27BonOpenRouterinput $0.2596 → $0.285/1M, output $2.385 → $2.4/1M↑+10%
- MiniMax M2onOpenRouteroutput $1 → $1.02/1M↑+2%
GPU rental price changes
GPU rental prices(44)
- A100 SXM 80GBonTensorDockon-demand $1.8 → $1.42/hr↓−21%
- B200onTogether AIon-demand $9.95 → $8.19/hr↓−18%
- A40onVast.aion-demand $0.32 → $0.29/hr↓−9%
- RTX 3090onVast.aion-demand $0.14 → $0.13/hr↓−7%
- L4onVast.aion-demand $0.34 → $0.32/hr↓−6%
- B200onRunPodon-demand $5.98 → $5.89/hr↓−2%
- RTX 6000 AdaonRunPodon-demand $0.74 → $0.77/hr↑+4%
- RTX 4090onTensorDockon-demand $0.35 → $0.37/hr↑+6%
- A100 SXM 80GBonRunPodon-demand $1.39 → $1.49/hr↑+7%
- GB300 SXM6onVerdaon-demand $7.99 → $8.62/hr↑+8%
Most active providers
- RunPod12 changes
- Verda10 changes
- OpenRouter9 changes
- Vast.ai7 changes
- Oblivus4 changes
- Crusoe3 changes
- Vultr3 changes
- TensorDock2 changes
- Together AI2 changes
- Latitude.sh1 change
Using these numbers
You're welcome to quote this report. Please link to it as the source:
Source: Inference Hub, AI API & GPU Price Report, July 2026 (https://inferencehub.org/price-report/2026-07)
How we collect prices: from each provider's own pricing page or API, checked weekly; see our methodology. Percentages compare the new price with the old; when input and output prices both moved, the larger move is shown. Get daily alerts by RSS or on Telegram.