Qwen3-235B: Pricing, Context, and Benchmarks
Qwen3-235B is Alibaba's Qwen3 model released on 2025-04-29. It runs on a 128K-token context window, priced at $0.50/M input and $2.00/M output. General-purpose Qwen3 with hybrid thinking. Open weights available under Apache 2.0.
Input
$0.50/M
Output
$2.00/M
Context
128K
Max output
8K
Specifications
| Provider | Alibaba |
|---|---|
| Family | Qwen3 |
| Released | 2025-04-29 |
| Status | active |
| Context window | 128K tokens |
| Max output | 8K tokens |
| Modalities | text |
| Capabilities | function calling json mode streaming thinking tool use |
| Tokenizer | qwen |
| API endpoint | https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions |
| Documentation | Official docs ↗ |
| Last updated | 2026-05-29 |
Cost calculator
For 1,000 tokens of input and 1,000 tokens of output per call, this model costs $0.0025 per call. For typical usage:
| Workload | Tokens / call (in/out) | $/call | $/1K calls | $/1M calls |
|---|---|---|---|---|
| Short Q&A | 200 / 100 | $0.000300 | $0.30 | $300 |
| Standard chat | 1K / 500 | $0.001500 | $1.50 | $1,500 |
| RAG with retrieval | 4K / 500 | $0.003000 | $3.00 | $3,000 |
| Long doc summary | 20K / 1K | $0.012000 | $12.00 | $12,000 |
| Long context (100K input) | 100K / 2K | $0.053000 | $53.00 | $53,000 |
For more precise estimates including caching discounts, use the LLM API Cost Calculator.
Other Alibaba models
Common comparisons
- Qwen3-235B vs DeepSeek-R1 Side-by-side specs, pricing, and benchmarks
- Qwen3-235B vs Llama 3.3 70B Side-by-side specs, pricing, and benchmarks
- Qwen3-235B vs Gemini 2.5 Flash Side-by-side specs, pricing, and benchmarks
- Qwen3-235B vs DeepSeek-V3 Side-by-side specs, pricing, and benchmarks
API setup
curl https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions \
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "qwen3-235b", "messages": [{"role": "user", "content": "Hello"}]}'See the official documentation for the full request/response schema. For pre-flight token counting before paying, use the appropriate token counter.