Alibaba Qwen · LLM API pricing

Qwen3.6 27B API pricing: cheapest provider

4 providers sell Qwen3.6 27B. SiliconFlow is cheapest; the most expensive charges 1.3× as much for the same model.

Cheapest$1.03per 1M · SiliconFlow
Input · output$0.30 / $3.20per 1M tokens
Price gap1.3×across 4 providers
Accuracy85.9%GPQA
Context262Kmax tokens

$1.03 SiliconFlow1.3× spread$1.35 Novita AI

Compare all 4 providers ⇄ Compare with another model

Already paying for Qwen3.6 27B? Check if you overpay →

Standard list prices in USD from each provider's pricing page · last price change Sep 28, 2026 · How we collect and rank prices

Ranked for your volume

Qwen3.6 27B monthly cost by provider

Change the monthly tokens and the ranking updates instantly, using each provider's standard list price.

Input tokens per month (millions)Prompts, context and documents you send.Output tokens per month (millions)Tokens the model generates.

More requirements: caching, context, features, provider

Cached share of input (%)0 = no caching. Only documented cache-read rates qualify.Required context per request (tokens)0 = any. Providers without a sufficient known limit are excluded.Required featureProvider requirement

Provider	Estimated / month	Input / 1M	Output / 1M	Context
1 SiliconFlowInference provider - lowest for your volumePricing page	$9.40	$0.30	$3.20	262,144
2 DeepInfraInference providerPricing page	$9.60	$0.32	$3.20	262,144
3 Alibaba Cloud Model StudioModel developerSingapore (International) list price; other regions differ.Pricing page	$13.20	$0.60	$3.60	262,144
4 Novita AIInference providerPricing page	$13.20	$0.60	$3.60	262,144

Order depends on listed price for your volume only. No provider pays for its position.

Input volume includes the cached share, not an additional token volume. This is a list-price estimate, not a quote: cache-write charges, cache eligibility/TTL, batch discounts, long-context surcharges, regional pricing and taxes are excluded. Check the provider's terms for your workload before switching.

Share this comparison as a Markdown table

For a post, a doc or a forum answer. The source line links back here, so readers can check today's prices.

Markdown

| Qwen3.6 27B provider | Input / 1M tokens | Output / 1M tokens | 10M in + 2M out per month |
|---|---:|---:|---:|
| SiliconFlow | $0.30 | $3.20 | $9.40 |
| DeepInfra | $0.32 | $3.20 | $9.60 |
| Alibaba Cloud Model Studio | $0.60 | $3.60 | $13.20 |
| Novita AI | $0.60 | $3.60 | $13.20 |

Standard list prices as of Oct 31, 2026. Source: [AISelect.online](https://aiselect.online/models/qwen3-6-27b)

Accuracy: GPQA Diamond (graduate-level science questions) measured by Epoch AI (CC BY 4.0) on qwen3.6-27b, run Aug 7, 2026. The model is the same at every provider, though quantized deployments can score lower.

Price history

The cheapest price has not changed since tracking started on Sep 28, 2026: $1.03 per 1M tokens at SiliconFlow. All price changes

Switch to the cheapest provider

Switch to SiliconFlow

Already using the OpenAI SDK? Change the base URL, key and model name:

from openai import OpenAI
import os

client = OpenAI(
    base_url="https://api.siliconflow.com/v1",
    api_key=os.environ["SILICONFLOW_API_KEY"],
)
response = client.chat.completions.create(
    model="Qwen/Qwen3.6-27B",
    messages=[{"role": "user", "content": "Hello"}],
)

Base URL from SiliconFlow's API docs, model name from the open pricing datasets. Check the model page in SiliconFlow's docs before you switch production traffic.

Compare Qwen3.6 27B with another model

Or pick any model

Model A

Model B

Before you pick a provider

SiliconFlow lists Qwen3.6 27B 24% below the developer's own API (Alibaba Cloud Model Studio) on a blended basis. Check rate limits, data retention and model quantization before switching.

Keep comparing

More Alibaba Qwen models

Qwen3 32B, Qwen3.8 27B, Qwen2.5-72B-Instruct, Qwen3 235B A22B Instruct 2507, Qwen3.6 35B-A3B, Qwen2.5 Coder 32B Instruct, Qwen3.5 397B-A17B, Qwen3.8 2.4T A95B

All Alibaba Qwen API prices - All LLM API prices - Watch Qwen3.6 27B prices (RSS) - Methodology