Qwen3 235B A22B Instruct 2507
by Alibaba (Qwen) · qwen/qwen3-235b-a22b-2507 · #66 cheapest of 325 paid models
Prices updated Jul 22, 2026, 1:35 AM UTC · refreshed hourly
Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...
Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol
Pricing
Input
$0.09
per 1M tokens
Output
$0.55
per 1M tokens
Blended (3:1)
$0.205
3 input : 1 output
Cache read
—
per 1M cached tokens
Cache write
—
per 1M tokens
Cache & batch economics
Effective prices for Qwen3 235B A22B Instruct 2507 (no cache-read rate published — hits billed as normal input).
| Cache hit rate | Effective blended $/1M | Example request* | vs no cache |
|---|---|---|---|
| 0% | $0.205 | $0.00973 | — |
| 50% | $0.205 | $0.00973 | −0% |
| 90% | $0.205 | $0.00973 | −0% |
*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates (available for OpenAI / Anthropic / Google — not flagged for this creator). Confirm on the vendor’s pricing page.
What a request costs
| Workload | Input tokens | Output tokens | Cost / request | Cost / 1K requests |
|---|---|---|---|---|
| Short chat message | 500 | 300 | $0.00021 | $0.21 |
| Document summary | 8,000 | 1,000 | $0.00127 | $1.27 |
| Codebase question (RAG) | 30,000 | 2,000 | $0.0038 | $3.80 |
| Long-context analysis | 150,000 | 5,000 | $0.0163 | $16.25 |
Performance
Intelligence Index
18.2
AA composite quality
Coding Index
—
Math Index
71.7
Agentic Index
—
Output speed
72 tok/s
median
Time to first token
1.10s
median
Time to first answer
1.10s
after reasoning tokens
Artificial Analysis benchmarks
| Benchmark | Score |
|---|---|
| MMLU-Pro | 82.8% |
| GPQA Diamond | 75.3% |
| Humanity's Last Exam | 10.6% |
| LiveCodeBench | 52.4% |
| SciCode | 36.0% |
| MATH-500 | 98.0% |
| AIME | 71.7% |
| AIME 2025 | 71.7% |
| IFBench | 46.1% |
| AA-LCR (Long Context Reasoning) | 31.2% |
| Terminal-Bench Hard | 15.2% |
| τ²-Bench (Telecom) | 33.3% |
Design Arena Elo
| Category | Elo |
|---|---|
| Website | 1,085 |
| UI components | 1,009 |
| Data viz | 1,095 |
| 3D | 1,053 |
| Game dev | 1,010 |
| Code categories | 1,072 |
Specs
Context window
262K
tokens
Max output
16K
tokens
Input modalities
text
Tokenizer
Qwen3