impulse

Qwen3.8 27B

Open weights

Qwen3.8 27B is an open-weight dense vision-language model from Qwen.

API id Qwen3.8-27BReleased 14 Aug 2026 (Aug 2026)
Input / 1M
$0.40
at DeepInfra
Output / 1M
$3.00
at DeepInfra
Context window
262K
tokens
Max output
262K
tokens per reply
Hosts
10
10 listings
Released
Aug 2026
14 Aug 2026

Pricing

List prices at DeepInfra, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.40$3.00
Cached input
90% off input
$0.04

What it costs a month

At DeepInfra list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$313
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$214
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$358
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$1,680
Use your own numbers

Where to run it

10 hosts serve Qwen3.8 27B. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
qwen3-8-27b
$0.17 / $0.50
−76% vs headline
qwen3.8-27b
$0.18 / $1.28
−57% vs headline
qwen3.8-27b
$0.383 / $2.73
−8% vs headline
ai&Cloud
qwen3.8-27b
$0.40 / $3.00
DeepInfraInference
Qwen3.8-27B
$0.40 / $3.00
W&B InferenceInference
Qwen3.8-27B
$0.40 / $3.00
Qwen3.8-27B
$0.45 / $3.00
+4% vs headline
qwen3.8-27b
$0.50 / $3.00
+7% vs headline
GMI CloudInference
Qwen3.8-27B
$0.45 / $3.20
+8% vs headline
GroqInference
qwen3.8-27b
$0.80 / $4.00
+52% vs headline

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
TextImages
Produces
Text
Features
Prompt caching
Open weights
License: Apache-2.0
Runs in about 66.7 GB of GPU memory
Runtimes: Ollama, SGLang, llama.cpp, vLLM

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

  • EmpirioLabs AI
    qwen3-8-27b
    $0.17 / $0.50
    76% less
  • CLF AI Gateway
    qwen3.8-27b
    $0.18 / $1.28
    57% less
  • Kyma API
    qwen3.8-27b
    $0.383 / $2.73
    8% less

Smaller Alibaba Qwen models

No cheaper Alibaba Qwen model keeps its key capabilities.

Other frontier labs

  • Gemini 3.5 Flash Lite
    Google · on Google AI Studio
    $0.30 / $2.50
    Has every key capability
    19% less
  • GPT-5.4 Nano
    OpenAI · on OpenAI
    $0.20 / $1.25
    Has every key capability
    56% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    81% less
  • GPT-6 Luna
    OpenAI · on OpenAI
    $0.10 / $0.50
    Has every key capability
    81% less

Open-weight picks

  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    35% less
  • Inkling Small
    Independent · on DeepInfra
    $0.45 / $1.20
    Has every key capability
    39% less
  • DeepSeek Flash
    DeepSeek · on DeepSeek
    $0.30 / $1.20
    Has every key capability
    50% less
  • Qwen3.7 Plus
    Alibaba Qwen · on Replicate
    $0.276 / $1.10
    Has every key capability
    54% less

Model line

Newer and older versions, at each lab's own price.

  1. Aug 2026Qwen3.8 27BLatest$0.40 / $3.00
  2. Aug 2026Qwen 3.8 27B$0.99 / $1.49
  3. Apr 2026Qwen3.6 27B$0.422 / $2.53
  4. Feb 2026Qwen3.5 27B$0.30 / $2.40

Call it

Using DeepInfra's API.

Python
from litellm import completion

response = completion(
    model="deepinfra/Qwen3.8-27B",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Qwen3.8 27B cost?

On DeepInfra, Qwen3.8 27B costs $0.40 per million input tokens and $3.00 per million output tokens. Cached input costs $0.04 per million tokens. The cheapest host is EmpirioLabs AI, at $0.17 per million input tokens and $0.50 per million output tokens.

Which APIs offer Qwen3.8 27B?

Qwen3.8 27B runs on 10 APIs: EmpirioLabs AI, CLF AI Gateway, Kyma API, ai&, DeepInfra, W&B Inference, Nebius AI Studio, StreamLake, GMI Cloud and Groq.

What is Qwen3.8 27B's context window?

262K tokens, with up to 262K tokens in one reply.

Is Qwen3.8 27B being retired?

No retirement date is scheduled for it on any API we track.

Is there a newer version of Qwen3.8 27B?

No. Qwen3.8 27B is the newest model in its line.

What are cheaper alternatives to Qwen3.8 27B?

Models that cost less and keep most of what it can do: Gemini 3.5 Flash Lite on Google AI Studio (19% less), GPT-5.4 Nano on OpenAI (56% less) and Kimi K2.7 Code on CLF AI Gateway (35% less). Qwen3.8 27B itself costs 76% less on EmpirioLabs AI. Savings are on a mix of 3 input tokens to 1 output token.

When was Qwen3.8 27B released?

Alibaba Qwen released it on 14 Aug 2026.

What is Qwen3.8 27B's API model id?

Use Qwen3.8-27B with DeepInfra. Other hosts use their own ids: qwen3-8-27b on EmpirioLabs AI, qwen3.8-27b on CLF AI Gateway, qwen3.8-27b on Kyma API, qwen3.8-27b on ai&.

Sources

Raw catalog entry for Qwen3.8-27B on DeepInfra
{
  "max_output_tokens": 262144,
  "max_input_tokens": 262144,
  "input_cost_per_token": 4e-7,
  "output_cost_per_token": 0.000003,
  "cache_read_input_token_cost": 4e-8,
  "supports_prompt_caching": true,
  "litellm_provider": "deepinfra",
  "mode": "chat",
  "supports_function_calling": true,
  "supports_tool_choice": true,
  "supports_response_schema": true,
  "supports_reasoning": true,
  "supports_vision": true,
  "source": "https://deepinfra.com/pricing"
}