AI Cost Index logo AI Cost Index LIVE
JA EN

Model A

Gemma 7B

Groq
VS

Model B

GPT-OSS Safeguard 20B

AWS Bedrock ConverseGroq
USD / 1M tokens Cheaper: Gemma 7B
Real-world costs vary by usage mix, cache behavior, and pricing mode.

Quick Take

Updated: 2026-06-03

Gemma 7B is cheaper overall here, and the larger gap appears on output pricing rather than input pricing.

  • Gemma 7B has the lower visible input price in the current snapshot.
  • Gemma 7B has the lower visible output price in the current snapshot.
  • Cached pricing is limited or incomplete for at least one side.

GPT-OSS Safeguard 20B show additional pricing modes in the current snapshot. Final billing still needs to be checked against the official pricing page and the exact mode you plan to use.

Comparison Table

Unified display rule: USD / 1M tokens

Input / 1M
$0.0500Standard
$0.0700Standard
Output / 1M
$0.0800Standard
$0.2000Standard
Cached input / 1M
N/ACached
N/ACached
Cached output / 1M
N/ACached
N/ACached
Advanced pricing
NTier/Mode
YTier/Mode
Other Pricing Modes
N
Y
Last check
2026-03-10
2026-06-03
Official Price

Premium Report

Need Deeper Pricing Insight?

Get cross-vendor cost intelligence, including hidden token/API costs,
plus full-period change history in CSV format.

Cross-Vendor Benchmark Historical CSV Instant Download
Buy via PayPal - Last Month Report (CSV) — $1,500

Instant download