Search Rankings

This page displays Arena AI A daily snapshot of the top 30 models on the leaderboard, making it easy to quickly understand recent model performance. For the full leaderboard, filters, and the latest changes, please visit the Arena AI official website.

This leaderboard evaluates model performance on web search / information retrieval tasks.

Data updated at: 2026-09-01 13:25:31 UTC / 2026-09-01 21:25:31 CST (Beijing time)

The leaderboard reflects the preferences of specific evaluations and user votes, and does not mean a model will definitely be better for your task. When choosing a model, also consider price, speed, context length, tool calling, privacy, and regional availability.

Top 30

Rank	Rank range	Model	Score	Votes	Price $/million tokens	Context
1	1-2	gpt-5.6-sol-xhigh 1	1257 (±7)	29,663	$4 / $20	1.1M
2	1-2	claude-opus-4-6-search 1	1253 (±5)	134,699	$5 / $25	1M
3	3-5	gpt-5.5-search 1	1242 (±5)	89,873	$5 / $30	1.1M
4	3-6	claude-opus-4-7 1	1233 (±5)	91,394	$5 / $25	1M
5	3-7	claude-fable-5 1	1230 (±8)	41,795	$10 / $50	1M
6	4-8	ernie-5.1 1	1227 (±10)	3,788	N/A	N/A
7	5-8	claude-sonnet-4-6-search 1	1221 (±5)	134,905	$1.50 / $7.50	1M
8	6-13	grok-4.5 1	1213 (±7)	31,505	$2 / $6	500K
9	8-13	gemini-3.1-pro-grounding 1	1210 (±5)	113,282	N/A	N/A
10	8-16	gemini-3-pro-grounding 1	1207 (±6)	37,024	$2 / $12	N/A
11	8-16	gpt-5.2-search 1	1207 (±6)	52,712	$0.88 / $7	400K
12	8-17	claude-opus-4-8 1	1204 (±6)	70,998	$5 / $25	1M
13	8-17	grok-4.20-multi-agent-beta-0309 1	1204 (±5)	109,553	$1.25 / $2.50	1M
14	10-18	gpt-5.1-search 1	1199 (±5)	59,909	$0.63 / $5	400K
15	10-18	gemini-3-flash-grounding 1	1198 (±5)	149,334	N/A	N/A
16	10-18	gpt-5.4-search 1	1197 (±5)	110,116	$2.50 / $15	1.1M
17	12-18	claude-sonnet-5-search 1	1194 (±7)	40,230	$2 / $10	1M
18	14-19	grok-4.20-beta1 1	1189 (±6)	53,921	N/A	N/A
19	18-22	claude-opus-4-5-search 1	1180 (±6)	61,573	$5 / $25	200K
20	19-23	gpt-5.2-search-non-reasoning 1	1172 (±6)	75,658	$0.88 / $7	400K
21	19-23	grok-4-1-fast-search 1	1171 (±5)	81,507	$1.25 / $2.50	2M
22	19-23	grok-4-fast-search 1	1171 (±5)	41,794	$0.20 / $0.50	2M
23	20-24	grok-4.3 1	1165 (±5)	91,483	$1.25 / $2.50	1M
24	23-24	claude-sonnet-4-5-search 1	1158 (±5)	127,378	$1.50 / $7.50	1M
25	25-29	claude-opus-4-1-search 1	1148 (±5)	76,933	$15 / $75	200K
26	25-29	o3-search 1	1144 (±5)	20,644	$1 / $4	200K
27	25-30	gemini-2.5-pro-grounding 1	1142 (±5)	83,404	$0.63 / $5	1M
28	25-31	grok-4-search 1	1142 (±6)	19,108	$3 / $15	N/A
29	25-31	ppl-sonar-reasoning-pro-high 1	1139 (±6)	29,055	$1 / $1	127.1K
30	27-32	gpt-5-search 1	1133 (±6)	20,781	$0.63 / $5	400K

How to read this table

• Rank / rank range: Relative ranking estimated by Arena AI based on head-to-head votes; the rank range indicates rank fluctuations within the confidence interval.
• Score: Relative score, suitable for comparing models on the leaderboard at the same point in time.
• Votes / sessions: Sample size reference; when the sample size is small, rankings are usually more likely to fluctuate.
• Price $/million tokens: Reference price per million input / output tokens.
• Context: The maximum context length supported by the model.

Three things to confirm again when choosing a model

1. Whether the provider actually offers this model, and whether it is available in your region and account;
2. Whether the API price, rate limits, and context length suit your task;
3. Run a small-scale test with 3 to 5 real tasks; do not rely only on the overall leaderboard rank.

Data source

Data from Arena AI official search leaderboard, updated daily by GitHub Actions. For model pricing, licenses, and capabilities, refer to the official information from the model provider.