Gemini 3.1 Pro Preview
GoogleProprietaryReleased Feb 19, 2026
On the arena.ai Agents leaderboard, Gemini 3.1 Pro Preview has a task success rate of -3.69% (—, rank #35, 95% interval -5.01%–-2.37%). Gemini 3.1 Pro Preview ranks #19 on the LMArena Text leaderboard (style control) with an Arena score of 1486.8 (±3.0, 129,523 votes). Artificial Analysis Intelligence Index: 29.7. API price: $2.00 input / $12.00 output per 1M tokens. Context: 1.05M. Nearest rivals: Kimi K3 (1488.1) above and Gemini 3 Pro (1485.5) below.
Quick answer
- Agents leaderboard
- #35 of 45 · -3.69% success
- Text (LMArena)
- #19 · 1486.8
- 17NAS overall
- #35 of 86
- API price (in / out)
- $2.00 / $12.00 per 1M tokens
- Closest rival
- Kimi K3 · 1488.1
- Cheaper alternative
- Gemini 3.8 Flash: Text 1496.3, output $3.75 (69% cheaper)
- Data updated
- Oct 22, 2026 · synced daily
Sources: arena.ai, LMArena, Artificial Analysis, OpenRouter
Key metrics
Models ranked nearby
Sorted by Arena score| # | Model | Arena | Intel. | Output $ | tok/s |
|---|---|---|---|---|---|
| 13 |  Muse Spark | 1489.0 | 31.3 | – | – |
| 14 | Gemini 3.7 Flash | 1488.9 | 39.1 | $3.75 | – |
| 15 |  Kimi K3 | 1488.1 | 43.6 | $15.00 | 39 |
| 16 | Gemini 3.1 Pro PreviewThis page | 1486.8 | 29.7 | $12.00 | 130 |
| 17 | Gemini 3 Pro | 1485.5 | 28.0 | $12.00 | – |
| 18 |  GPT-5.6 Sol | 1484.5 | 47.0 | $20.00 | – |
| 19 | Gemini 3.6 Flash | 1483.0 | 34.0 | $3.75 | – |
More from Google
| # | Model | Arena | Intel. | Output $ | tok/s |
|---|---|---|---|---|---|
| 1 | Gemini 4 Argon | 1523.9 | 52.6 | $10.00 | 80 |
| 8 | Gemini 3.8 Flash | 1496.3 | 40.9 | $3.75 | 144 |
| 14 | Gemini 3.7 Flash | 1488.9 | 39.1 | $3.75 | – |
| 17 | Gemini 3 Pro | 1485.5 | 28.0 | $12.00 | – |
| 19 | Gemini 3.6 Flash | 1483.0 | 34.0 | $3.75 | – |
| 26 | Gemini 3.5 Flash | 1477.5 | 32.6 | $9.00 | – |
Pricing
- Official API list price: $2.00 input and $12.00 output per 1M tokens (USD).
- At a typical 3:1 input-to-output ratio, the blended cost is about $4.50 per 1M tokens.
- Among 81 models with an output price, it is the #59 cheapest.
- The cheapest OpenRouter route charges $2.00 input / $12.00 output.
Sources & methodology
On the arena.ai Agents leaderboard, Gemini 3.1 Pro Preview has a task success rate of -3.69% (—, rank #35, 95% interval -5.01%–-2.37%).
- arena.ai Agents — agent task success rate, 95% interval, sessions and rank (via the public Hugging Face dataset, CC BY 4.0).
- LMArena — Arena score (Text leaderboard with style control, the same view arena.ai shows by default), 95% confidence interval, votes and rank.
- Artificial Analysis — Intelligence / Coding Index, output speed and latency (medians).
- OpenRouter — Route prices, context length and listing date.
Arena scores use the Bradley-Terry model (not Elo); ± is the half-width of the 95% confidence interval. Latency figures are Artificial Analysis medians; for reasoning models, time to first token and to first answer include the thinking phase. Data syncs daily at 14:20 (UTC+8).
Limits: Arena scores come from anonymous users voting on chats, so they reflect conversational preference rather than agent or coding benchmarks; with few votes the interval is wide and ranks move. Prices and speeds change as providers update them.
FAQ
How does Gemini 3.1 Pro Preview do on the Agents leaderboard?
On the arena.ai Agents leaderboard, Gemini 3.1 Pro Preview has a task success rate of -3.69% (—, rank #35, 95% interval -5.01%–-2.37%).
What rank is Gemini 3.1 Pro Preview right now?
Gemini 3.1 Pro Preview is #16 on our leaderboard of 69 scored models (Arena score 1486.8); its raw LMArena rank is #19.
What does Gemini 3.1 Pro Preview cost via API?
Input $2.00 / output $12.00 per 1M tokens (official API list price; the OpenRouter quote when no official price is available). At a typical 3:1 input-to-output ratio, the blended cost is about $4.50 per 1M tokens.
How fast is Gemini 3.1 Pro Preview?
Artificial Analysis measures about 130 output tokens per second, 18 s to the first token and 18 s to the first answer token; it is a reasoning model, so latency includes thinking time.
How close is Gemini 3.1 Pro Preview to the models ranked next to it?
It is 1.3 points behind Kimi K3 (1488.1) and 1.3 points ahead of Gemini 3 Pro (1485.5); gaps smaller than the confidence interval (±3.0) are not significant.
Is Gemini 3.1 Pro Preview open source? Can it be self-hosted?
Gemini 3.1 Pro Preview is a closed model, available only via official APIs or hosted services.
What is the context length of Gemini 3.1 Pro Preview?
1.05M (1,048,576 tokens, as listed on OpenRouter).
Gemini 3.1 Pro Preview vs Kimi K3: which is better?
On the LMArena Text leaderboard Gemini 3.1 Pro Preview scores 1486.8 and Kimi K3 scores 1488.1 (a 1.3-point gap, within the confidence interval, so not significant). Agent task success: Gemini 3.1 Pro Preview -3.69%, Kimi K3 1.77%. On price, Gemini 3.1 Pro Preview costs $12.00 and Kimi K3 $15.00 per 1M output tokens.
Is there a cheaper model close to Gemini 3.1 Pro Preview?
Yes: Gemini 3.8 Flash scores 1496.3 on the Text leaderboard (Gemini 3.1 Pro Preview: 1486.8) and costs $3.75 per 1M output tokens, about 69% less than Gemini 3.1 Pro Preview ($12.00).
Where does the Gemini 3.1 Pro Preview data come from and how often is it updated?
Agent success rates come from the arena.ai Agents leaderboard (CC BY 4.0), Text scores from LMArena, intelligence and speed from Artificial Analysis, and prices and context from official list prices and OpenRouter. 17NAS syncs daily at 06:20 UTC; this page was last updated Oct 22, 2026.