Gemini 3.1 Pro
AI models review
As of Aug 12, 2026, Gemini 3.1 Pro is rated 1486 Elo on LM Arena (94,728 Arena votes).
Benchmark and price facts
- Vendor
- License
- proprietary
- LM Arena text
- 1486 Elo (#13 of 20)
- LM Arena coding
- 1447 Elo (#44)
- Price (input)
- $2 / 1M tokens
- Price (output)
- $12 / 1M tokens
- Context window
- 1,048,576 tokens
- AA Intelligence Index
- 47.7 / 100
- AA Coding Index
- 68.8 / 100
- Output speed (AA)
- 126 tokens/s
- Human votes
- 94,728
Elo and votes from LM Arena; intelligence, coding and speed indices from Artificial Analysis; price and context window from OpenRouter. Real fetched values, attributed, never blended.
Is Gemini 3.1 Pro worth it in 2026?
Gemini 3.1 Pro sits at rank 13 of 20 in LM Arena text with an Elo of 1486, backed by about 94,728 human votes, giving it a reasonable but mid-tier standing in general chat. Its coding Elo of 1447 places it at rank 44, which is notably weak on that dimension. The Artificial Analysis Intelligence Index is 47.7 out of 100 and the Coding Index is 68.8 out of 100. Output speed is 126 tokens per second and the context window is a very large 1,048,576 tokens. Pricing is 2 USD per 1M input and 12 USD per 1M output, which is on the higher end for a model at this performance tier.
Long-document work where the massive 1,048,576-token context window is the priority, and general conversational use where mid-tier text quality is acceptable.
The coding rank of 44 is poor relative to its text rank, and the output price of 12 USD per 1M tokens is high for a model ranked 13 of 20 in text quality, making it a costly choice with limited coding strength.
AI-generated editorial synthesis of the source scores and sampled review themes, not a hands-on test. Reviewer experiences and allegations are not independently verified facts.
Rated near Gemini 3.1 Pro
Closest scores in ai models, on the same source.