Qwen3 32B
alibaba-qwen3-32b
Qwen instruction model for multilingual chat, reasoning, and tool use
Specification
most-agreed values
Context
33K
Max output
33K
Released
2025-04-30
Knowledge cutoff
—
Retires
—
Open weights
yes
Input
text
Output
text
Price
US dollars per million tokens · most-agreed
Input
$0.25
Output
$0.55
Quality
4 benchmarks · 2 sources
Composite
41st
percentile of 334 scored models
Rank
197 / 334
±0.08 sd
Evidence
4 × 2
3 or more benchmarks, from 2 or more sources
Effort range
—
one configuration only
coding33rd1/3 bench
reasoning38th2/4 bench
math44th1/7 bench
Every score
one row per source and configuration — nothing averaged away
| Benchmark | Score | Configuration | Source | Run |
|---|---|---|---|---|
| GPQA diamond | 59.9 ±2.8 | default · 2 runs | Epoch AI | 2026-08-30 |
| OTIS Mock AIME 2024-2025 | 45.0 ±5.1 | default · 2 runs | Epoch AI | 2026-08-30 |
| Aider polyglot | 40.0 | default · aider diff | Aider | 2025-05-08 |
| Chess Puzzles | 3.0 ±1.0 | default · 2 runs | Epoch AI | 2026-08-28 |
The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.
Available from 2 hosts
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| digitalocean | $0.25 | $0.55 | — | — | 33K | 33K | — |
| gradient_ai · gradient-ai | — | — | — | — | 131K | 41K | — |
Other listings of this model
this page is one of them
The corpus files Qwen3 32B under several keys. This page is the alibaba-qwen3-32b listing. The full record — every host, every price and every benchmark score — is on the main Qwen3 32B page.