Quick answers

Index above 55 and the answer starts within two seconds.

As of 25 Oct 2026, Gemma 4 26B A4B by Google leads this list (0.5 s), ahead of Claude Sonnet 4.6 (0.7 s).

8 models qualify · data as of 25 Oct 2026

1. 1Gemma 4 26B A4B0.50 s
2. 2Claude Sonnet 4.60.71 s
3. 3Qwen3.5 Plus (thinking)1.01 s
4. 4Claude Sonnet 4.5 (high)1.11 s
5. 5Claude Opus 4.6 (high)1.22 s
6. 6Claude Opus 4.71.22 s
7. 7GPT-5 nano1.25 s
8. 8Muse Spark 1.11.55 s

SearchAll providersAlibabaAnthropicGoogleMetaOpenAIOpen weights8 of 8

#	Model	First answer ↓	Index ↕	Price $/M ↕	Speed tok/s ↕
1	Muse Spark 1.1Meta	1.55 s	62.5	$2.00	200
2	GPT-5 nanoOpenAI	1.25 s	55.9	$0.138	82
3	Claude Opus 4.7Anthropic	1.22 s	63.2	$10.00	60
4	Claude Opus 4.6highAnthropic	1.22 s	62.7	$10.00	39
5	Claude Sonnet 4.5highAnthropic	1.11 s	58.1	$6.00	38
6	Qwen3.5 PlusthinkingAlibaba	1.01 s	57.3	$0.258	54
7	Claude Sonnet 4.6Anthropic	0.71 s	57.1	$6.00	44
8	Gemma 4 26B A4BGoogleopen ↗	0.50 s	56.2	$0.198	46