智商指数
86.5 #37
AAQI 86.0
输出速度
260 t/s
ITL 3.8 ms/tok
混合价格 (3:1)
$0.01 /1M
输入 $0.007/M · 输出 $0.015/M
上下文
64K
≈ 160 页 A4 · 最大输出 8,192
Quality Domain Radar
6-Axis Frontier Evaluation vs Category Average & Flagship
Deepgram Flux
Claude Mythos Preview (#1)
Average (68)
缺少直接基准测量时,雷达图各轴由类别得分近似得出 — 详见 评分方法。
优势与取舍
优势所在
- Fastest conversational STT — 260 t/s-class realtime factor
- Enterprise accuracy on calls and meetings
- Purpose-built turn-taking for voice agents
需要注意的权衡
- Audio input only — pair with an LLM for reasoning
- Noisy environments need endpoint tuning
杀手级用例
Voice agents
Streaming STT with turn detection — the front end of every serious voice agent.
Call-center analytics
Real-time transcription and intent extraction.
Live captions
Broadcast and meeting captioning.
开发者快速上手与 API
请先设置
DEEPGRAM_API_KEY — 获取 API 密钥curl -X POST "https://api.deepgram.com/v1/listen?model=flux-general-en" \
-H "Authorization: Token $DEEPGRAM_API_KEY" \
-H "Content-Type: audio/wav" \
--data-binary @call.wavimport os, requests
with open("call.wav", "rb") as f:
resp = requests.post(
"https://api.deepgram.com/v1/listen",
params={"model": "flux-general-en"},
headers={"Authorization": f"Token {os.environ['DEEPGRAM_API_KEY']}"},
data=f,
)
print(resp.json()["results"]["channels"][0]["alternatives"][0]["transcript"])const audio = await fs.promises.readFile("call.wav");
const res = await fetch("https://api.deepgram.com/v1/listen?model=flux-general-en", {
method: "POST",
headers: { Authorization: `Token ${process.env.DEEPGRAM_API_KEY}`, "Content-Type": "audio/wav" },
body: audio,
});
console.log((await res.json()).results.channels[0].alternatives[0].transcript);Streaming STT — post audio bytes, get a transcript. Flux adds acoustic turn detection for voice agents.
官方资源与链接
API 部署
对比托管商与延迟
Groq、Cerebras、Together、DeepInfra、Fireworks、Bedrock 与 Azure 的吞吐量、TTFT 和缓存定价。
GroqCerebrasTogether AIDeepInfraFireworks AIAWS BedrockAzure AI
各榜单得分
分类能力得分
基准测试分项
实时规格与架构
底层解析
Flux — conversational speech-recognition transformer
- Turn detection from acoustic + textual cues — knows when a speaker finished
- Streaming-first: partial transcripts in real time
- Flux TTS sibling enables full voice-agent loops
- Multilingual Flux variant available
发布日期
2026年9月28日
架构
Flux — conversational speech-recognition transformer
参数量
Not disclosed
开源协议
Proprietary
上下文窗口
64,000 tokens
最大输出
8,192 tokens
模态
audio
缓存价格
读 $0.002 · 写 $0.01 /1M
推理
—
输入价格
$0.007/M
输出价格
$0.015/M
输出速度
260 t/s
首Token延迟
—
数据更新
2026年9月30日
每月账单预估
每月预估支出
$0.00
正在计算相对 GPT-4o 的节省额…
引用与嵌入
GitHub Markdown 徽章
Loading badge markdown...BibTeX
@misc{llmpodium2026deepgramflux,
title = {Benchmark Analysis and Intelligence Evaluation of Deepgram Flux},
author = {LLMPodium Research},
year = {2026},
url = {https://llmpodium.com/models/deepgram-flux}
}嵌入评分徽章
将实时 Podium 评分徽章添加到 GitHub 或 Hugging Face README —— 链接回实时排名。
预览
Markdown
Loading badge snippet...HTML
Loading badge snippet...更多 Independent 模型
替代模型推荐
基于能力、定价和基准测试得分与 Deepgram Flux 最接近的替代模型。