Qwen: Qwen3 235B A22B Instruct 2507
by QwenQwen: Qwen3 235B A22B Instruct 2507 is a large language model from Qwen. It costs $0.090 per million input tokens and $0.550 per million output tokens. Its context window is 262K tokens.
- Input / 1M tokens
- $0.090
- Output / 1M tokens
- $0.550
- Cached input / 1M
- -
- Context window
- 262K
Not supported
tokens
Who serves it cheapest
6 hosts serve Qwen3 235B A22B Instruct 2507. Same weights, same API - the price difference is pure margin and routing.
| Provider | Input / 1M | Output / 1M | Context | Throughput | Uptime 24h |
|---|---|---|---|---|---|
| GMICloudCheapestfp8 | $0.087 | $0.350 | 262K | - | 0.0% |
| DeepInfrafp8 | $0.090 | $0.550 | 262K | - | 62.6% |
| Venicefp8 | $0.150 | $0.750 | 128K | - | 72.0% |
| Nebiusfp8 | $0.200 | $0.600 | 262K | - | 70.9% |
| Parasailfp8 | $0.140 | $0.800 | 131K | - | 100.0% |
| $0.220 | $0.880 | 262K | - | 100.0% |
The spread between GMICloud and Google is 2.5× for identical weights. Quantization and context limits differ, so check both columns before switching.
Benchmarks
Independent scores published alongside the catalogue.
- Intelligence index
- 12.0
About Qwen3 235B A22B Instruct 2507
Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...
Specifications
| Model ID | qwen/qwen3-235b-a22b-2507 |
|---|---|
| Provider | Qwen |
| Context window | 262K tokens |
| Max output | 16K tokens |
| Input modalities | text |
| Output modalities | text |
| Knowledge cutoff | 2025-06-30 |
| Open weights | Yes - Qwen/Qwen3-235B-A22B-Instruct-2507 |
| Released | July 21, 2025 |
Cheaper alternatives
Models that cost less than Qwen3 235B A22B Instruct 2507 while keeping at least half its context window and every input modality it supports.
Claude Haiku 5.5
$0.100 in · $0.500 out
about the same cheaperGPT-6 Luna Pro
$0.100 in · $0.500 out
about the same cheaperGPT-6 Luna
$0.100 in · $0.500 out
about the same cheaperNemotron 3.5 Content Safety
$0.200 in · $0.200 out
about the same cheaperMinistral 3 14B 2512
$0.200 in · $0.200 out
about the same cheaperFrequently asked
How much does Qwen: Qwen3 235B A22B Instruct 2507 cost?
$0.090 per million input tokens and $0.550 per million output tokens.
What is the context window of Qwen: Qwen3 235B A22B Instruct 2507?
262K tokens, with up to 16K tokens of output per request.
Which provider serves Qwen: Qwen3 235B A22B Instruct 2507 cheapest?
GMICloud at $0.087 per million input tokens - 2.5× cheaper than Google, the most expensive of the 6 hosts serving it.
Confirm against the source: Qwen official pricing.