Claude Fable 5 versus Qwen3.8 Max 0902

What are we comparing?

This comparison pits two frontier models, the ones you pick when quality comes first against each other: Claude Fable 5 and Qwen3.8 Max 0902, one from Anthropic, the other from Alibaba Cloud (Qwen). The first dates from June 9, 2026, the second from September 2, 2026.

All figures below come from public sources checked on October 22, 2026: models.dev and OpenRouter for prices and capabilities, Epoch AI for scores.

Performance

On our capability index, Claude Fable 5 scores 90.8 out of 100 and Qwen3.8 Max 0902 88.4. Claude Fable 5 leads by 2.4 points.

On the 3 benchmarks reported for both, Claude Fable 5 wins 1 time and Qwen3.8 Max 0902 1:

• GPQA Diamond (PhD-level science questions): Claude Fable 5 85.9%, Qwen3.8 Max 0902 92.3%
• OTIS Mock AIME (competition mathematics problems): Claude Fable 5 100%, Qwen3.8 Max 0902 100%
• FrontierMath (very hard research-level mathematics): Claude Fable 5 87%, Qwen3.8 Max 0902 65.6%

Which one costs less?

Claude Fable 5 costs $10 per million input tokens and $50 per million output tokens; Qwen3.8 Max 0902 costs $2 and $6.

For a typical agent task (read 20,000 tokens, write 2,000), expect $0.30 with Claude Fable 5 versus $0.052 with Qwen3.8 Max 0902. Qwen3.8 Max 0902 is 5.8 times cheaper, a gap that adds up quickly over 1,000 tasks: $248 difference.

Run the numbers with your own volumes in the agent cost calculator.

Context and capabilities

Both models have the same context window: 1,000,000 tokens, roughly 1,500 pages.

Reasoning mode: Claude Fable 5 yes, Qwen3.8 Max 0902 yes. Tool calling: Claude Fable 5 yes, Qwen3.8 Max 0902 yes. Image input: Claude Fable 5 yes, Qwen3.8 Max 0902 yes.

Neither publishes its weights: both are used through an API.

In short

Our reading of the data:

• Choose Claude Fable 5 for a higher capability index.
• Choose Qwen3.8 Max 0902 for a lower cost and a more recent release.

This verdict relies on public figures; it does not replace a trial on your own tasks.