Qwen3 Coder 30B A3B Instruct
by Alibaba (Qwen) · qwen/qwen3-coder-30b-a3b-instruct · #34 cheapest of 325 paid models
Prices updated Jul 22, 2026, 2:18 AM UTC · refreshed hourly
Qwen3-Coder-30B-A3B-Instruct is a 30.5B parameter Mixture-of-Experts (MoE) model with 128 experts (8 active per forward pass), designed for advanced code generation, repository-scale understanding, and agentic tool use. Built on the...
Compare vs:Claude Sonnet 5Claude Opus 4.6Claude Haiku 4.5GPT-5.6 Sol
Pricing
Input
$0.07
per 1M tokens
Output
$0.27
per 1M tokens
Blended (3:1)
$0.12
3 input : 1 output
Cache read
—
per 1M cached tokens
Cache write
—
per 1M tokens
Cache & batch economics
Effective prices for Qwen3 Coder 30B A3B Instruct (no cache-read rate published — hits billed as normal input).
| Cache hit rate | Effective blended $/1M | Example request* | vs no cache |
|---|---|---|---|
| 0% | $0.12 | $0.00741 | — |
| 50% | $0.12 | $0.00741 | −0% |
| 90% | $0.12 | $0.00741 | −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 | $0.000116 | $0.116 |
| Document summary | 8,000 | 1,000 | $0.00083 | $0.83 |
| Codebase question (RAG) | 30,000 | 2,000 | $0.00264 | $2.64 |
| Long-context analysis | 150,000 | 5,000 | $0.0119 | $11.85 |
Performance
Intelligence Index
13.6
AA composite quality
Coding Index
—
Math Index
29
Agentic Index
—
Output speed
102 tok/s
median
Time to first token
1.53s
median
Time to first answer
1.53s
after reasoning tokens
Artificial Analysis benchmarks
| Benchmark | Score |
|---|---|
| MMLU-Pro | 70.6% |
| GPQA Diamond | 51.6% |
| Humanity's Last Exam | 4.0% |
| LiveCodeBench | 40.3% |
| SciCode | 27.8% |
| MATH-500 | 89.3% |
| AIME | 29.7% |
| AIME 2025 | 29.0% |
| IFBench | 32.7% |
| AA-LCR (Long Context Reasoning) | 29.0% |
| Terminal-Bench Hard | 15.2% |
| τ²-Bench (Telecom) | 34.5% |
Design Arena Elo
| Category | Elo |
|---|---|
| Website | 1,115 |
| UI components | 1,089 |
| Data viz | 1,121 |
Specs
Context window
262K
tokens
Max output
33K
tokens
Input modalities
text
Tokenizer
Qwen3