Qwen3 235B A22B Instruct is Alibaba's language model with a 262K context window and up to 33K output tokens, available from 10 providers, starting at $0.090 / 1M input and $0.580 / 1M output. An instruction-tuned update of the Qwen3 235B A22B MoE model with significant improvements in instruction following, logical reasoning, and general capabilities.
Specifications
Canonical IDalibaba-qwen3-235b-a22b-instruct
TypeLanguage
StatusActive
CreatorAlibabaAlibaba
Providers
Context Window262K tokens
Max Output33K tokens
Input ModalitiesText
Output ModalitiesText
Parameters235B
HuggingFace Likes773
HuggingFace Downloads (30d)150,781
HuggingFace Downloads (all-time)1,182,969
Benchmarks
Intelligence Index
17.0
#259
Coding Index
14.0
#238
Math Index
23.7
#194
MMLU-Pro
0.8
#150
GPQA
0.6
#246
HLE
0.0
#308
LiveCodeBench
0.3
#190
AIME
0.3
#70
IFBench
0.4
#262
Time to First Token
1.30s
#367
SciCode
0.3
#223
MATH-500
0.9
#63
AIME 2025
0.2
#194
LCR
0.0
#335
TerminalBench Hard
0.1
#233
TAU2
0.3
#242
Output TPS
64.0
#180

Capabilities

Input1/5
Text
Image·
Audio·
Video·
PDF·
Output1/5
Text
Image·
Audio·
Video·
Embedding·
Capabilities3/13
Reasoning·
Adaptive Reasoning·
Function Calling
Parallel Function Calling
Structured Outputs
Native JSON Schema·
Web Search·
URL Context·
Computer Use·
Code Execution·
File Search·
Prompt Caching·
Assistant Prefill·

Pricing by Provider

Cost Calculator

Preset:

Versions

VersionReleasedContextInput / 1MOutput / 1MStatus
MiroThinker 1.7Available
Qwen3 30B A3B Thinking262K$0.080$0.400Available
Qwen3 Coder 30B A3B Instruct262K$0.070$0.260Available
Qwen3 30B A3B Instruct262K$0.090$0.300Available
Qwen3 235B A22B Thinking262K$0.149$0.880Deprecated
Qwen3 Coder1.0M$0.220$1.80Available
Qwen3 30B A3B131K$0.080$0.290Deprecating
Qwen3 235B A22B262K$0.071$0.100Available
Qwen3 235B A22B Instruct262K$0.090$0.580Current
Qwen3 Coder 480B A35B Instruct262K$0.220$1.30Available
KwaiPilot KAT Coder262K$0.900$0.900Available

Model IDs