Qwen
Alibaba's open-weight family, with a coder model in nearly every size.
Qwen is Alibaba's model family and the largest open-weight line there is — dozens of sizes from a few billion parameters up to the Max models, most of them published so you can run them yourself. For vibe coders the draw is the Qwen Coder models: strong at code for the price, cheap to call through a provider, and small enough at the low end to run on your own machine. The flagship Max models take text, images and video and go up to a million tokens of context.
At a glance — Qwen3.8 27B
- Cost
- $0.42 in · $3.00 out per 1M tokensBudgetFree version too
- Memory
- 1M tokens — about 1,500 pages
- Reads
- Text · Images · Video
- Can do
- Use tools (agents)
- Structured output
- Thinking mode
- Good at
- Coding: better than 71% · Agentic work: better than 85% · Overall: better than 61%
- Reliability
- 99.94% uptime, last 24h
- Released
- Aug 14, 2026
Advanced details
- API model ID
qwen/qwen3.8-27b- Free version
qwen/qwen3.8-27b:free- Max output
- 131K tokens
- Cached input
- $0.085 per 1M
- Tokenizer
- Qwen
- Coding index
- 68.1
- Agentic work index
- 45.8
- Overall index
- 33.7
Providers
The same model, hosted by different companies. OpenRouter picks one per request and falls back to the next if it fails.
| Provider | Input /1M | Output /1M | Cached /1M | Max output | Uptime 24h |
|---|---|---|---|---|---|
| Reka | $0.09 | $4.40 | $0.085 | 131K | 99.12% |
| Wafer | $0.091 | $4.40 | $0.085 | 236K | 99.79% |
| DekaLLM | $0.096 | $4.40 | $0.087 | 236K | 99.82% |
| Darkbloom · fp4 | $0.10 | $1.80 | — | 33K | 99.70% |
| Ionstream · fp8 | $0.105 | $2.55 | $0.10 | 66K | 98.65% |
| DeepInfra · bf16 | $0.15 | $1.875 | $0.038 | 236K | 97.89% |
| Phala | $0.199 | $2.075 | $0.042 | 236K | 99.28% |
| Mancer 2 · fp8 | $0.20 | $2.50 | — | 236K | 99.70% |
| Parasail · fp8 | $0.24 | $2.20 | $0.05 | 236K | 99.94% |
| Chutes · fp8 | $0.24 | $2.20 | $0.024 | 66K | 99.09% |
| AkashML · fp8 | $0.25 | $2.20 | $0.05 | 131K | 99.89% |
| CoreWeave · fp8 | $0.40 | $3.00 | $0.15 | 236K | 99.90% |
| Novita | $0.42 | $3.00 | $0.085 | 131K | 99.81% |
| Alibaba | $0.425 | $2.55 | $0.085 | 131K | 99.42% |
| Cloudflare | $0.45 | $3.20 | $0.05 | 236K | 90.79% |
| Venice · fp8 | $0.45 | $3.20 | — | 66K | 99.43% |
Supported parameters
- frequency_penalty
- include_reasoning
- logit_bias
- logprobs
- max_tokens
- min_p
- presence_penalty
- reasoning
- reasoning_effort
- repetition_penalty
- response_format
- seed
- stop
- structured_outputs
- temperature
- tool_choice
- tools
- top_a
- top_k
- top_logprobs
- top_p
Live from OpenRouter, refreshed hourly. Scores from Artificial Analysis.
Key info
- Pricing
- Open source
- Category
- Models
- Best for
- Intermediate
- Made by
- Alibaba
- Weights
- Open, most sizes
- Good for
- Cheap coding models, and running one yourself