Inference
EN
Alibaba

Qwen3 32B

Qwen3 32B is a standard model from Alibaba. It has 10 scored results in 4 categories. It has no index.

  • Open weights
  • Standard
  • Input: text
  • Released 2025-04
UnrankedEnough results for an index.
IndexAll categories, out of 100.UnrankedMedian 51.0
CoverageIndex weight that has results.35%
SpeedOutput tokens per second.46/sMedian 67/s
First tokenSeconds to the first output token.0.4 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.08Median $0.60
Output / 1MUS dollars per 1M output tokens.$0.28Median $2.55
ContextMaximum tokens in one request.131KMedian 262K
EloLMArena rating and rank.1340±9 #219Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
qwen/qwen3-32b
Max outputMaximum output tokens per request.
16K
Knowledge cutoffLast date of training data.
31 Mar 2025
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
—
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
4 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Deep InfraHost$0.08$0.2841K
OpenRouterRouter$0.08$0.28131K
SiliconFlowHost$0.14$0.57131K
Amazon BedrockCloud$0.15$0.6033K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.None counted——
CodingCode writing and repair.None counted——
ReasoningLogic problems and puzzles.None counted——
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.4 counted32.6—
MultilingualTasks in many languages.1 counted34.8—
InstructionTasks with strict prompt rules.1 counted30.7—
MathMath problems.4 counted21.8—

Results

10 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
GPQA diamond · Epoch AIKnowledge54.1% ±2.8137 / 217
Instruction-Following EvalInstruction83.2%46 / 70
OTIS Mock AIME 2024-2025Math23.1% ±5.1142 / 200
MMLU-ReduxKnowledge85.7%16 / 19
Graduate-Level Google-Proof Q&AKnowledge54.6%120 / 141
MATH-500 Problem SetMath88.6%24 / 27
INCLUDEMultilingual70.9%11 / 12
American Invitational Mathematics Examination 2024Math31.0%25 / 27
MMMLUKnowledge76.5%14 / 15
American Invitational Mathematics Examination 2025Math20.2%52 / 53
C-EvalKnowledge83.3%12 / 12
Berkeley Function Calling Leaderboard v3Agentic63.0%—
LiveCodeBench v5Coding31.3%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseEpoch AICC BYmodels.devMIT

History

5 releases
Output / 1M$0.28 → $0.28
23 Oct 202612 Nov 2026

Questions

How much does Qwen3 32B cost?
$0.08 per 1M input tokens, $0.28 per 1M output tokens.
How fast is Qwen3 32B?
46 output tokens per second, 0.4 s to the first token.
What is the context window of Qwen3 32B?
131K tokens.
Is Qwen3 32B open source?
Yes. Its weights are published.
When was Qwen3 32B released?
April 2025.

More from Alibaba

All models
Qwen3 235B-A22B38.1Qwen3 235B-A22B (Reasoning)UnrankedQwen3 32B (Reasoning)UnrankedQwen3 30B A3BUnrankedQwen3 8BUnranked