BroadRouter
Get started

DeepSeek V4.1 Flash API

Updated

DeepSeek V4.1 Flash runs on BroadNet Inference as broadnet/deepseek-v4.1-flash through the OpenAI-compatible https://inference.broadnet.ai/v1. It costs $0.175 input, $0.0055 cached input and $0.52 output per 1M tokens, or from $60 a month per session on a flat plan (one request at a time), with a 1,048,576-token context. It needs a qualifying top-up (not the starting credit), and inference content is not retained.

At a glance

  • Text
  • Vision (image input)
  • Reasoning
  • 1,048,576-token context on every request
  • Text and image input
  • Reasons before answering; reasoning counts toward max_tokens
  • Unlimited tokens on the Full Moon, Orbit and Meteor session plans, one request at a time per session
Price and limits
Model IDbroadnet/deepseek-v4.1-flash
Endpoint/v1/chat/completions
Supported inputstext, image (vision: understanding, not generation)
Context1,048,576 tokens, prompt and answer together
Input$0.175 per 1M tokens
Cached input$0.0055 per 1M tokens (reported cache hits only)
Output$0.52 per 1M tokens
Pay-as-you-go limits30 requests a minute, 4 in flight per key
AccessTop-up required: opens after a qualifying top-up; the starting credit does not cover it
Session plansFrom $60 a month per session, one request at a time per session (plans)

Prices on this page are from BroadNet's price feed snapshot dated , not checked live here. See official pricing for current rates.

Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).

What does BroadNet publish about DeepSeek V4.1 Flash?

From BroadNet's Terms of Service, checked 15 November 2026.

What BroadNet publishes about this model
HostingPartner-hosted: BroadNet's Terms of Service (§7.3) say "DeepSeek V4.1 Flash remains open-weight when partner-hosted"
Data handlingPartner-served models run under zero-data-retention terms with BroadNet's US inference partners (privacy statement)
Extra rulesUse of the hosted service also follows the partner restrictions in BroadNet's Acceptable Use Policy

How BroadNet chooses and serves models: its GPU hardware and serving approach and the models it evaluated and ruled out.

Published benchmark results

BroadNet did not run these benchmarks; results are as published by the DeepSeek model card and Epoch AI and may use different settings. BroadRouter is BroadNet's site. They refer to the published model deepseek-ai/DeepSeek-V4.1-Flash; not measured on BroadNet's serving.

Published benchmark results for DeepSeek V4.1 Flash, with sources
BenchmarkResultSetting or noteSource
GPQA Diamond (Pass@1)90.9Instruct model, maximum reasoning effort (reasoning_effort=100)DeepSeek model card, checked 2026-11-16
HLE (Pass@1)36.8 (39.1†)Instruct model, maximum reasoning effort (reasoning_effort=100); the figure in brackets is the text-only subsetDeepSeek model card, checked 2026-11-16
Terminal-Bench 2.1 (Pass@1)90.6Instruct model, maximum reasoning effort (reasoning_effort=100); DeepSeek Harness, Minimal modeDeepSeek model card, checked 2026-11-16
DeepSWE v1.1 (Resolved)74.2Instruct model, maximum reasoning effort (reasoning_effort=100); mini-SWE harnessDeepSeek model card, checked 2026-11-16
Codeforces (Rating)3471Instruct model, maximum reasoning effort (reasoning_effort=100)DeepSeek model card, checked 2026-11-16
GPQA Diamond0.898Max reasoning setting; Epoch run deepseek-v4.1-flash_max, mean score on a 0 to 1 scaleEpoch AI, checked 2026-11-16
SimpleQA Verified0.429Max reasoning setting; Epoch run deepseek-v4.1-flash_max, mean score on a 0 to 1 scaleEpoch AI, checked 2026-11-16
OTIS Mock AIME 2024-20250.983Max reasoning setting; Epoch run deepseek-v4.1-flash_max, mean score on a 0 to 1 scaleEpoch AI, checked 2026-11-16

Epoch AI rows: Epoch AI, Capabilities & benchmarking, licensed CC BY 4.0, shown as the mean score on a 0 to 1 scale, rounded to three decimals (hover a value for the exact figure in Epoch's data file).

Side by side: DeepSeek V4.1 Flash vs GLM 5.3, DeepSeek V4.1 Flash vs Qwen3.8 27B and DeepSeek V4.1 Flash vs BroadNet Gemma 4 VL 26B, with prices, context and workload costs. DeepSeek V4.1 Flash is also sold by 16 other providers, from $0.14 input and $0.42 output per 1M tokens on their own pricing pages: DeepSeek V4.1 Flash API pricing by provider. The same text is a different number of tokens on each model: see how DeepSeek V4.1 Flash's tokenizer compares on Arabic, English and code in the tokenizer efficiency dataset.

Call it in Python

Install the OpenAI SDK with pip install openai, export BROADNET_API_KEY, then run:

Python
from openai import OpenAI
import base64, os

client = OpenAI(base_url="https://inference.broadnet.ai/v1", api_key=os.environ["BROADNET_API_KEY"])
image = base64.b64encode(open("invoice.jpg", "rb").read()).decode()

response = client.chat.completions.create(
    model="broadnet/deepseek-v4.1-flash",
    max_tokens=2000,
    messages=[{"role": "user", "content": [
        {"type": "text", "text": "What is the total on this invoice?"},
        {"type": "image_url", "image_url": {"url": "data:image/jpeg;base64," + image}},
    ]}],
)
print(response.choices[0].message.content)

Or with curl

curl
curl https://inference.broadnet.ai/v1/chat/completions \
  -H "Authorization: Bearer $BROADNET_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"broadnet/deepseek-v4.1-flash","messages":[{"role":"user","content":"Reply with one short sentence."}]}'

DeepSeek V4.1 Flash reasons before it answers. The reasoning counts toward max_tokens and is reported as usage.completion_tokens_details.reasoning_tokens, inside the completion (output) token count, so budget for it as output. Leave 1,000 tokens or more, or set reasoning_effort to low, or the answer can come back empty. Why answers come back empty.

Run it on a session plan

The plans differ only in how long they last. Each session runs one request at a time; add a session for each agent working in parallel. Work out sessions and the charge with the session planner.

Full Moon

Billed per month

$60 a month per session

One request at a time per session

Founders offer: 25% off for the first 3 months, $45 a month per session, then the standard $60. BroadNet's feed says: “Enter a founders code at checkout.” No code is published here.

Start Full Moon on inference.broadnet.ai

Orbit

Billed per week

$15 a week per session

One request at a time per session

Buy a week on inference.broadnet.ai

Meteor

Billed per hour

$1 an hour per session

One request at a time per session

Start an hour on inference.broadnet.ai

All plans cover DeepSeek V4.1 Flash only, with no token meter for the purchased period. One session runs one request at a time, so parallel agents need one session each. Context limits and the Acceptable Use Policy still apply; images are included. Session pricing does not guarantee throughput or uptime. Prices in US dollars; BroadNet's plans page says tax is added at checkout where it applies.

Guides for DeepSeek V4.1 Flash

Use DeepSeek V4.1 Flash in the console

The BroadNet console on inference.broadnet.ai manages accounts and keys. It does not receive your choice from this page, so:

  1. Sign in or create an account.
  2. Add a qualifying top-up (the starting credit does not cover this model), then create a pay-as-you-go key. Or buy a session plan and create a plan key: see plans.
  3. Send this model ID as model to /v1/chat/completions: broadnet/deepseek-v4.1-flash

Open the BroadNet consoleEstimate the costRead the API reference

Compare DeepSeek V4.1 Flash

Questions

What is the model ID for DeepSeek V4.1 Flash?

Pass broadnet/deepseek-v4.1-flash as model to /v1/chat/completions. The base URL is https://inference.broadnet.ai/v1.

How much does DeepSeek V4.1 Flash cost?

$0.175 input, $0.0055 cached input and $0.52 output per 1M tokens, from your account balance. It is also the model on the session plans: $60 a month per session with no token meter, one request at a time per session.

What is the context window of DeepSeek V4.1 Flash?

1,048,576 tokens per request, prompt and answer together.

Can I use DeepSeek V4.1 Flash without a top-up?

No. DeepSeek V4.1 Flash opens after a qualifying top-up. The starting credit covers BroadNet Gemma 4 VL 26B, BroadNet Transcribe Arabic, BroadNet Transcribe and Gemma hourly keys.

How does DeepSeek V4.1 Flash score on benchmarks?

As published, the DeepSeek model card reports 90.9 on GPQA Diamond (Pass@1) and 90.6 on Terminal-Bench 2.1 (Pass@1); Epoch AI measured 0.898 on GPQA Diamond in its own run, as a mean score on a 0 to 1 scale (checked 2026-11-16). BroadNet did not run these benchmarks, and sources may use different settings, so the figures are not directly comparable. Every published result we found, with its setting and source.

Is my data stored?

Inference content is not retained: prompts, images and answers are processed for the request only. Necessary account, billing and security metadata, such as timings, token counts and cost, is retained as described in BroadNet's privacy statement. Read how.

Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity