Qwen3 235B A22B 2507 Instruct
Intelligence
25
#207/521
Coding
22.1
#194/427
Math
71.7
#92/265
Speed
66 tok/s
TTFT: 1.05s
Pricing
$0.70 / $2.80
per 1M tokens (in/out)
Qwen3 235B A22B 2507 Instruct is Alibaba’s AI model designed for various applications in natural language processing. It processes at a speed of 66.275 tokens per second and is priced at $0.7 per million input tokens, targeting professional users.
When to Use Qwen3 235B A22B 2507 Instruct
✓ Best For
- Natural language understanding tasks.
- Coding assistance and support.
- Mathematical problem-solving applications.
✗ Not Ideal For
- Tasks requiring high-level creativity.
- Real-time interactive applications due to TTFT of 1.048s.
How Qwen3 235B A22B 2507 Instruct Compares
Intelligence Index · Higher is better
xAIByteDance SeedAlibaba
Benchmark Profile
Coding Index
KimiGoogleAlibabaxAIServiceNow
Output Speed · tok/s
XiaomiNVIDIAAlibabaAnthropic
Math Index
AlibabaInclusionAI
Intelligence · Coding · Math
Intelligence
Coding
Math
All Benchmark Scores (39)
| Benchmark | Score | |
|---|---|---|
| Intelligence Index | 25 | |
| Coding Index | 22.1 | |
| Math Index | 71.7 | |
| MMLU-Pro | 828% | |
| GPQA | 753% | |
| LiveCodeBench | 524% | |
| HLE | 106% | |
| SciCode | 36% | |
| IFBench | 46.1% | |
| LCR | 31.2% | |
| TerminalBench Hard | 15.2% | |
| Tau2 | 33.3% | |
| AIME | 71.7% | |
| AIME 2025 | 71.7% | |
| MATH 500 | 98% | |
| Aider-Polyglot | 57.3% | |
| ARC-AGI | 41.8% | |
| Arena-Hard v2 | 0.8 | |
| BFCL-v3 | 70.9% | |
| Creative Writing v3 | 87.5% | |
| CSimpleQA | 84.3% | |
| HMMT25 | 55.4% | |
| IFEval | 88.7% | |
| Include | 79.5% | |
| LiveBench 20241125 | 75.4% | |
| LiveCodeBench v6 | 51.8% | |
| MMLU-ProX | 79.4% | |
| MMLU-Redux | 93.1% | |
| Multi-IF | 77.5% | |
| MultiPL-E | 87.9% | |
| PolyMATH | 50.2% | |
| SimpleQA | 54.3% | |
| SuperGPQA | 62.6% | |
| Tau2 Airline | 44% | |
| Tau2 Retail | 71.3% | |
| WritingBench | 85.2% | |
| ZebraLogic | 95% | |
| Arena Chat | 8.9 | |
| Arena Coding | 1.5 |
Data: Artificial Analysis · Updated: April 10, 2026
Frequently Asked Questions (15)
When was Qwen3 235B A22B 2507 Instruct released?
Qwen3 235B A22B 2507 Instruct was released on July 21, 2025.
Who created Qwen3 235B A22B 2507 Instruct?
Qwen3 235B A22B 2507 Instruct was created by Alibaba.
How intelligent is Qwen3 235B A22B 2507 Instruct?
Qwen3 235B A22B 2507 Instruct scores 25 on the Artificial Analysis Intelligence Index, placing it above average among other open weight non-reasoning models of similar size (median: 20).
How fast is Qwen3 235B A22B 2507 Instruct?
Qwen3 235B A22B 2507 Instruct generates output at 54.6 tokens per second (based on Alibaba's API), which is above average compared to other open weight non-reasoning models of similar size (median: 54.6 t/s).
What is the latency of Qwen3 235B A22B 2507 Instruct?
Qwen3 235B A22B 2507 Instruct has a time to first token (TTFT) of 2.41s (based on Alibaba's API), which is somewhat higher than average compared to other open weight non-reasoning models of similar size (median: 2.25s).
How much does Qwen3 235B A22B 2507 Instruct cost?
Qwen3 235B A22B 2507 Instruct costs $0.70 per 1M input tokens (somewhat higher than average, median: $0.60) and $2.80 per 1M output tokens (somewhat higher than average, median: $2.33), based on Alibaba's API.
What is Qwen3 235B A22B 2507 Instruct API pricing?
Qwen3 235B A22B 2507 Instruct costs $0.70 per 1M input tokens and $2.80 per 1M output tokens (based on Alibaba's API). For a blended rate (3:1 input to output ratio), this is $1.23 per 1M tokens. Pricing may vary by provider.
How verbose is Qwen3 235B A22B 2507 Instruct?
When evaluated on the Intelligence Index, Qwen3 235B A22B 2507 Instruct generated 15M output tokens, which is at the higher end compared to other open weight non-reasoning models of similar size (median: 9.1M).
Is Qwen3 235B A22B 2507 Instruct a reasoning model?
No, Qwen3 235B A22B 2507 Instruct is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.
What input modalities does Qwen3 235B A22B 2507 Instruct support?
Qwen3 235B A22B 2507 Instruct supports text input.
What output modalities does Qwen3 235B A22B 2507 Instruct support?
Qwen3 235B A22B 2507 Instruct supports text output.
Can Qwen3 235B A22B 2507 Instruct process images?
No, Qwen3 235B A22B 2507 Instruct does not support image input. It can only process text.
Is Qwen3 235B A22B 2507 Instruct multimodal?
No, Qwen3 235B A22B 2507 Instruct is not multimodal. It only supports text input.
What is the context window of Qwen3 235B A22B 2507 Instruct?
Qwen3 235B A22B 2507 Instruct has a context window of 260k tokens. This determines how much text and conversation history the model can process in a single request.
Is Qwen3 235B A22B 2507 Instruct open source?
Yes, Qwen3 235B A22B 2507 Instruct is open weights. The model weights are publicly available and can be downloaded for self-hosting.
