Qwen: Qwen3.5 27B

One API key. Live prices. Ready-to-run code.

Qwen3.5-27B is an open-weight vision-language model with a hybrid architecture combining Gated Delta Networks and Gated Attention.

Served on OneInfer by Novita. Use it with the same API key and pay-as-you-go wallet as every other model in the catalog.

ProviderNovitaQwen/Qwen3.5-27B

Call Qwen3.5 27B in 4 steps

Step 1 of 4

Create a key in the Console. The same key works for every model on OneInfer.

Input tokens
$0.300 / 1M tokens
Output tokens
$2.400 / 1M tokens
Context
262K
Input
text, image
Output
text
Tool calling
Supported

About this model

Qwen3.5-27B is an open-weight vision-language model with a hybrid architecture combining Gated Delta Networks and Gated Attention. It features a 262K native context length, unified vision-language foundation, and supports scalable reinforcement learning for robust agentic tasks

Input modalities

  • text

    Accepted as model input

  • image

    Accepted as model input

Providers

Available routing options for this model through OneInfer.

  • Available
    Input tokens
    $0.300 / 1M tokens
    Output tokens
    $2.400 / 1M tokens
    Routing
    OneInfer optimized

Pricing

Current OneInfer pricing for this model.

UsagePrice
Input tokens$0.300 / 1M tokens
Output tokens$2.400 / 1M tokens

Performance and benchmarks

Independent evaluations and live cluster operational metrics

We haven’t measured this model yet. Published academic results are under Academic evaluations.

API guide

API docs
  1. Generate your access token

    Use your API key to generate the JWT access token required by the Models API.

  2. Add the JWT token

    Copy the generated JWT token and replace JWT_TOKEN in the example below.

  3. Check your credits

    If your balance is too low, before calling the model.

  4. Run the API example

    Choose your preferred language, copy the example, and send your first model request.

    curl https://api.oneinfer.ai/v1/ula/chat/completions \
      -H "Authorization: Bearer $JWT_TOKEN" \
      -H "Content-Type: application/json" \
      -d '{
        "model": "Qwen/Qwen3.5-27B",
        "messages": [
          { "role": "user", "content": "Hello!" }
        ]
      }'

Community reviews and ratings

Visitor and developer evaluations of Qwen: Qwen3.5 27B's real-world intelligence, code accuracy, and speed

No ratings yet

Top community tags

Be the first to rate and tag this model's coding and reasoning strengths.

No community reviews yet

Be the first to share developer feedback on Qwen: Qwen3.5 27B's latency, code accuracy, and reliability in production.