Black Forest Labs
FLUX.2 Dev开源
Open-weight FLUX.2 tier for self-hosting, fine-tuning and LoRA workflows.
智商指数
90.4 #15
AAQI 90.0
输出速度
28 t/s
ITL 35.7 ms/tok
混合价格 (3:1)
$0.03 /1M
输入 $0.03/M · 输出 $0.03/M
上下文
128K
≈ 320 页 A4 · 最大输出 4,096
Quality Domain Radar
6-Axis Frontier Evaluation vs Category Average & Flagship
FLUX.2 Dev
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
- Near-Pro image quality you can self-host
- Fits an RTX 4090 at Q4 — no datacenter needed
- Thriving LoRA/fine-tune ecosystem
- ~3× cheaper than Pro on hosted APIs
需要注意的权衡
- Slightly below Pro on hard prompts
- Non-commercial terms on some deployments
- No 4K output in the open checkpoint
杀手级用例
Self-hosted generation
Own the image stack — no per-image API fees.
Fine-tuned brand styles
LoRA adapters for house styles and characters.
Offline pipelines
Air-gapped rendering for sensitive assets.
开发者快速上手与 API
请先设置
BFL_API_KEY — 获取 API 密钥curl -X POST "https://api.bfl.ml/v1/flux-2-dev" \
-H "x-key: $BFL_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "Isometric render of a neon data center, sharp typography",
"width": 1024,
"height": 1024
}'import os, requests
resp = requests.post(
"https://api.bfl.ml/v1/flux-2-dev",
headers={"x-key": os.environ["BFL_API_KEY"]},
json={
"prompt": "Isometric render of a neon data center, sharp typography",
"width": 1024,
"height": 1024,
},
)
print(resp.json()["polling_url"]) # poll until status == "Ready"const res = await fetch("https://api.bfl.ml/v1/flux-2-dev", {
method: "POST",
headers: { "x-key": process.env.BFL_API_KEY, "Content-Type": "application/json" },
body: JSON.stringify({
prompt: "Isometric render of a neon data center, sharp typography",
width: 1024,
height: 1024,
}),
});
console.log((await res.json()).polling_url); // poll until status === "Ready"Async image API — the response returns a polling_url; poll it until status is Ready.
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
实时对决
实时规格与架构
底层解析
FLUX.2 Dev — 32B open rectified-flow transformer
- Same architecture as Pro, released under open weights
- Runs on a single 24GB GPU at Q4
- LoRA / DreamBooth fine-tuning supported
- Non-commercial-friendly open license
发布日期
2026年10月14日
架构
FLUX.2 Dev — 32B open rectified-flow transformer
参数量
32B
开源协议
Open Weights
上下文窗口
128,000 tokens
最大输出
4,096 tokens
模态
text + image
缓存价格
读 $0.01 · 写 $0.03 /1M
推理
—
输入价格
$0.03/M
输出价格
$0.03/M
输出速度
28 t/s
首Token延迟
—
数据更新
2026年10月22日
硬件与本地推理
数据来源:Convly AI(CC BY 4.0) ↗VRAM(4-bit Q4)
20 GB
最低 GPU
RTX 4090 24GB / RTX 5090 32GB
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{llmpodium2026flux2dev,
title = {Benchmark Analysis and Intelligence Evaluation of FLUX.2 Dev},
author = {LLMPodium Research},
year = {2026},
url = {https://llmpodium.com/models/flux-2-dev}
}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 Black Forest Labs 模型
替代模型推荐
基于能力、定价和基准测试得分与 FLUX.2 Dev 最接近的替代模型。