Skip to content
trending

AI Model Atlas / Qwen (Alibaba)

Qwen3 235B A22B Thinking 2507

One million input tokens cost $0.45 and one million output tokens $3.50. Context window: 128k tokens, about 213 A4 pages.

Data from 12 November 2026 API identifier: qwen/qwen3-235b-a22b-thinking-2507

What it costs in practice

Pricing per million tokens doesn't mean anything to anyone, so I convert it to three tasks that people actually give to models. The usage assumptions are listed for each row so you can verify them.

TaskInputOutputCost
1500-word articlebrief plus finished text in English 1,500 tokens 2,200 tokens $0.0084
100 chat queries700 tokens per question and 400 per answer 70,000 tokens 40,000 tokens $0.17
Hour of work with an agentagent reads repository and writes fixes 250,000 tokens 30,000 tokens $0.22
The calculation does not include tax, intermediary markup, or context caching, which with repeated prompts can cut input costs by several dozen percent.

Technical details

ProviderQwen (Alibaba)
Modalitiestext → text
Context window128k tokens (~213 A4 pages)
Maximum response16k tokens
Step-by-step reasoningyes
Tool callingyes
Forced JSON formatno
Knowledge cutoff2025-06-30
Cache readno data
Content moderationno extra filter

How it compares to the rest

The database contains 468 models. In terms of input pricing Qwen3 235B A22B Thinking 2507 is more expensive than 48 percent of paid models. A longer context window is available in 400 models.

Input price history

2026-09-08 · $0.232026-11-12 · $0.45

The chart shows the price of a million input tokens on consecutive measurement days. History starts on the day I launched the atlas and grows with each day. I don't backfill data.

Alternatives at a similar price

Models from other providers whose cost for the same article comes out closest.

ModelInput / MOutput / M1500-word articleContext
Kimi K2.6Moonshot AI $0.65 $3.41 $0.0085 262k
Pareto 26.10Unbiased $0.80 $3.20 $0.0082 1M
Pareto 26.10 PreviewUnbiased $0.80 $3.20 $0.0082 1M
Nova Pro 1.0Amazon $0.80 $3.20 $0.0082 300k

Compare Qwen3 235B A22B Thinking 2507 with another model or go back to the full database 468 models.

// site map
Quantslant in numbers 1,435posts in archive32news from the last 7 days18video models in the observatory
🎬 AI Video Benchmarks: Higgsfield DoP vs OpenAI Sora • Higgsfield vs Runway Gen-3 • AI Video Credit Calculator
× ‹ enlarge ›