Use cases › Speed

Answer before the user notices: low latency and realtime voice

Voice agents, autocomplete and live UI, where the first token has to arrive in a few hundred milliseconds.

22 chat models with a measured first token at these constraints. Data observed 2026-10-26 07:36 UTC.

Showing models for: first token within 300 ms.

What it looks like in practice

DeepSeek-V4.1-Flash sends its first token in 114 ms; the eighth fastest chat model takes 207 ms.

Fastest live endpoint per model, time to first token. Teal rows are this page's picks.

Source: OpenRouter endpoint stats, one snapshot from one client location, observed 2026-10-26

Speed run: Qwen3.8-27B

median TTFT 11,694 ms · 93.3 tok/s

Receipt, parasail, 2026-10-09 · source

Start with these three

One pick per constraint group, each the leader on its group's rule: the cheapest endpoint that is fast enough, the fastest small-active MoE, the fastest dense model of 32B or fewer. The numbers and dates are the claim; the order is not a verdict.

Cheapest at speed

Qwen3-4B-Thinking-2507

Qwen

Cheapest input price among endpoints with a first token within 300 ms and at least 80 tokens per second: $0.010 per 1M (Nscale, 2026-10-26).

281 msfirst token

159 tok/soutput

$0.010per 1M input

Sources: huggingface.co 2026-10-26

Small active params

Qwen3-30B-A3B

Qwen

Fastest first token among mixture-of-experts chat models with 5B or fewer active parameters: 166 ms (DeepInfra, 2026-10-26).

166 msfirst token

3B of 30.5Bactive

$0.12per 1M input

Sources: huggingface.co 2026-10-26 · huggingface.co 2026-10-26

Dense, mid-size

Qwen3.8-27B

Qwen

Fastest first token among dense chat models of 32B parameters or fewer: 161 ms (Cerebras, 2026-10-26).

161 msfirst token

27.8Bparams

$0.020per 1M input

Sources: huggingface.co 2026-10-26 · huggingface.co 2026-10-26 · openrouter.ai 2026-10-26

Recipes

How people wire a model into this job. Notes are written by hand from the cited cards and docs; the model links go to the facts.

Prefill-only classification service

Model: Qwen/Qwen3-0.6B
Harness: vLLM with max_tokens=1 and logit readout over option tokens, or an encoder with a classification head
Weights: BF16 (fits edge/single-GPU class)

Score options from logits instead of generating text; no decode phase.
Batch aggressively; prefill is compute-bound and batches well.
Validate on a labeled set; small models drift with prompt wording.

Everything that fits 22

The first ten chat models with a measured first token, in the scoreboard's order for this job. The full list, sortable, lives in the scoreboard with these constraints.

Everything that fits: first 10 of 22
#
1	DeepSeek-V4.1-Flashdeepseek-ai	114 ms	283	$0.020	34	MIT
2	Qwen3.8-27BQwen	161 ms	887	$0.020	23	Apache 2.0
3	Qwen3-30B-A3BQwen	166 ms	113	$0.12	1	Apache 2.0
4	Qwen3-14BQwen	184 ms	91	$0.070	4	Apache 2.0
5	c4ai-command-r7b-12-2024CohereLabs	201 ms	106	—	1	CC BY-NC 4.0
6	Qwen3.5-397B-A17BQwen	204 ms	90	$0.39	13	Apache 2.0
7	GLM-5.2zai-org	205 ms	210	$0.032	27	MIT
8	Qwen3.5-27BQwen	207 ms	85	$0.20	7	Apache 2.0
9	Muse-Glimmer-30Bmeta-models	207 ms	99	$0.30	3	Apache 2.0
10	gpt-oss-120bopenai	218 ms	1095	$0.030	27	Apache 2.0

Show all 22 in the scoreboard →

Most-shared on this job

Links about these models ranked by how many newsletters and link sources shared them in the last 28 days; each source counts once.All links

Shared by 5
DeepSeek-V4.1-Flash-1017 (Hugging Face Repo)Hacker News, The Neuron, TLDR AI, AI Breakfast and 1 moreAbout DeepSeek-V4.1-Flash
huggingface.co · first seen 2026-10-17
Shared by 3
DeepSeek-V4.1-Flash-1017: Wish-Driven Programming Made RealThe Neuron, Turing Post, InterconnectsAbout DeepSeek-V4.1-Flash
api-docs.deepseek.com · first seen 2026-10-19
Shared by 2
DeepSeek V4.1 Flash: Large-Scale Inference DeploymentLatent Space / AINews, TLDR AIAbout DeepSeek-V4.1-Flash
zhuanlan.zhihu.com · first seen 2026-10-21
Shared by 2
Chinese AI lab DeepSeek updated its V4.1-Flash model, which it says now finishes tasks faster with shorter answers, and…The Rundown AI, Don't Worry About the VaseAbout DeepSeek-V4.1-Flash
x.com · first seen 2026-10-19
Shared by 2
DeepSeek V4.1 Flash 1017Hacker News, AlphaSignalAbout DeepSeek-V4.1-Flash
openrouter.ai · first seen 2026-10-17

How people judge this job

OpenRouter endpoint stats (first token, tokens per second)

None of these measure time to first audio for a voice agent. The first-token times here are text endpoints from OpenRouter's endpoint stats, one snapshot from one client location; a voice loop adds speech recognition and synthesis on top.

Nearest scoreboard view: Low latency.