Qwen3.8 27B is an open-weight dense vision-language model from Qwen.
Qwen3 32B
Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogu
SharkScore
#129 in Models
Why it scores 40.1
Qwen3 32B has a SharkScore of 40.1/100, ranked #129 of 608 in Models (updated 2026-11-19).
Each part compares Qwen3 32B with other listings in Models. Evidence covers 87% of the score's weight.
- Adoption
- 43 · 30%
- User reviews
- no data · 25%
- Maintenance
- 1 · 20%
- Trust & transparency
- 80 · 15%
- Momentum
- no data · 10%
Paying never changes this score. How ranking works · method v2026-11-18.2
RankShark review of Qwen3 32B
Automated evaluation by RankShark · 18 Nov 2026 · method v1.0 · How we review
We were not paid for this review.
As of 2026-11-18, Qwen3 32B is a 32.8B text model listed with a 131,072-token OpenRouter context and measured rates of $0.08/$0.28 per million input/output tokens [E1, E2, MEASURED].
Qwen3-32B is a 32.8B-parameter causal language model for text generation. The vendor describes thinking and non-thinking modes, and provides instructions for local use and API-compatible deployment [E1].
Best for
- People building text applications that need a choice between reasoning-oriented and general dialogue modes. · model card
- Developers building multilingual or tool-using applications. · model card
- Developers who want to deploy the model locally or through an OpenAI-compatible endpoint. · model card
Not for
- Teams that need 131,072 tokens of context without configuring long-context scaling; the vendor describes 32,768 tokens as native and YaRN scaling for longer contexts. · model card
Key features
- Thinking modes. The vendor says the model can switch between thinking and non-thinking modes, including through a setting in the chat template. · model card
- Long-context support. The vendor says the model natively supports 32,768 tokens and has been validated up to 131,072 tokens using YaRN scaling. · model card
- Tool use. The vendor says Qwen3 supports tool calling and recommends Qwen-Agent for agentic use. · model card
- Multilingual support. The vendor says the model supports more than 100 languages and dialects. · model card
- Deployment options. The vendor lists Transformers, SGLang, vLLM, Ollama, LMStudio, MLX-LM, llama.cpp, and KTransformers for local use or deployment. · model card
Pros
- The vendor describes both thinking and non-thinking modes, with a switch for choosing between them. · model card
- The vendor documents local deployment options and OpenAI-compatible API deployment. · model card
- The OpenRouter listing has a measured context window of 131,072 tokens. · OpenRouter
Cons
- The vendor says 131,072-token context requires YaRN scaling; it also warns static scaling can affect performance on shorter texts. · model card
- No free plan or free trial stated on the pages we checked.
- Whether customer data is used for model training is not stated on the pages we checked.
Pricing
| Plan | Price | Limits |
|---|---|---|
| OpenRouter listing | $0.08 per million input tokens; $0.28 per million output tokens | Token-based pricing. The OpenRouter listing reports a 131,072-token context window. |
Checked 18 Nov 2026
- Free plan
- Not stated (checked 18 Nov 2026)
- Free trial
- Not stated (checked 18 Nov 2026)
- Trains on customer data
- Not stated (checked 18 Nov 2026)
- Data retention
- Not stated (checked 18 Nov 2026)
- Data regions
- Not stated (checked 18 Nov 2026)
- Data processing agreement
- Not stated (checked 18 Nov 2026)
- AI models used
- Not stated (checked 18 Nov 2026)
- API
- OpenRouter lists chat completions, completions, messages, and responses endpoints. · OpenRouter
- SDKs
- Not stated (checked 18 Nov 2026)
- Webhooks
- Not stated (checked 18 Nov 2026)
- Platforms
- Hugging Face Transformers; SGLang; vLLM; Ollama; LMStudio; MLX-LM; llama.cpp; KTransformers. · model card
- Security claims
- No security certification stated (checked 18 Nov 2026)
What we measured
- Context window (OpenRouter listing) · now
- 131,072 tokens
- Price per million tokens (OpenRouter listing) · now
- $0.08 in, $0.28 out
- Package downloads · 30 days
- 3,694,930
How we tested · sources
- Hourly uptime and response-time checks of the product's URL.
- Repository, release and download data from GitHub, npm, PyPI and Hugging Face where they exist.
- We did not sign up, pay, or use the product as a customer.
- model card: https://huggingface.co/Qwen/Qwen3-32B · fetched 18 Nov 2026
- OpenRouter: https://openrouter.ai/qwen/qwen3-32b · fetched 18 Nov 2026
- 18 Nov 2026: Plans or prices changed on the pricing page.
- 18 Nov 2026: Plans or prices changed on the pricing page.
Something wrong? Report an error with a public link that shows it; corrections re-run the review.
Top alternatives to Qwen3 32B
See all- 67.1ModelOpen source
- 65.5
Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per t
ModelOpen source - 64.7
Qwen3.8 Flash is a multimodal reasoning model from Alibaba.
ModelOpen source - 64.3
Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026.
ModelOpen source