Compare AI models

Decision facts

• Gemma 3 4B is #255; Gemma 4 31B is #37 on the same current language board.
• Index scores are 26.0 for Gemma 3 4B and 71.9 for Gemma 4 31B. Their published uncertainty intervals do not overlap, but the aggregate Index does not predict performance on every task; compare the relevant benchmark configurations before choosing.
• Both positions use Measured evidence.
• Listed output price per 1M tokens is $0.10 for Gemma 3 4B and $0.40 for Gemma 4 31B. For 1,000 requests with 2,000 input and 500 output tokens each (2M input + 0.5M output), the listed-rate estimate is $0.15 for Gemma 3 4B versus $0.60 for Gemma 4 31B. Gemma 3 4B costs 75.0% less in this scenario. This excludes caching, batch discounts, prompt-length tiers, tool charges and retries; verify the selected endpoint before budgeting.
• Published context is 131,072 tokens for Gemma 3 4B and 262,144 for Gemma 4 31B.
• Weight access differs: Gemma 3 4B is gated; Gemma 4 31B is open.
• ModelCap currently lists 2 providers for Gemma 3 4B and 16 for Gemma 4 31B.

These are separate published fields, not a synthetic winner. ModelCap does not collapse price, access, context, and capability evidence into a hidden recommendation score.