Frontier model pricing · August 2026

Frontier intelligence.
A fraction of the price.

A visual comparison of list API pricing for Qwen 3.8 Max and the frontier model lineups from Anthropic and OpenAI, as published on vendor pricing pages. Qwen 3.8 Max undercuts every comparable frontier model on input, output, and cached-token pricing — while including a 1M-token context window at base price.

Qwen OpenAI Anthropic USD per 1M tokens Standard list pricing
7.5×
cheaper than Claude Fable 5
$8 vs $60 for 1M input + 1M output tokens at list price
4.4×
cheaper than GPT-5.6-sol
$8 vs $35 for the same 2M blended tokens, before any long-context surcharge
1M
token context window, included
Full 1M context at base price — no long-context surcharge (OpenAI bills up to 2× beyond short context)

List price, per million tokens

Lower is better

Input pricing

USD per 1M input tokens
Qwen 3.8 MaxBest value$2.00
GPT-5.6-sol$5.00
Claude Opus 5$5.00
Claude Fable 5$10.00
$0$5$10

Qwen 3.8 Max input is 60% below GPT-5.6-sol and 80% below Claude Fable 5 — the cheapest frontier-tier input price on the market.

Output pricing

USD per 1M output tokens
Qwen 3.8 MaxBest value$6.00
Claude Opus 5$25.00
GPT-5.6-sol$30.00
Claude Fable 5$50.00
$0$25$50

Output tokens: 80% below GPT-5.6-sol and 88% below Claude Fable 5 — the cheapest frontier-tier output price on the market.

The blended bottom line

1M input + 1M output tokens

Combined cost per 2M blended tokens

USD, list price, short context
Qwen 3.8 MaxBest value$8.00
Claude Opus 5$30.00
GPT-5.6-sol$35.00
Claude Fable 5$60.00
$0$30$60
The same 2M-token workload costs $8 on Qwen 3.8 Max — 4.4× cheaper than OpenAI’s flagship GPT-5.6-sol ($35) and 7.5× cheaper than Anthropic’s flagship Claude Fable 5 ($60).

Cache economics

Repeated-context workloads

Cached input (cache hits)

USD per 1M cached input tokens
Qwen 3.8 MaxBest value$0.25
GPT-5.6-sol$0.50
Claude Opus 5$0.50
Claude Fable 5$1.00
$0$0.50$1.00

Qwen cache hits ($0.25 implicit, $0.17 explicit) run 50–75% below frontier competitor cache pricing.

Cache writes

USD per 1M tokens written to cache
Qwen 3.8 MaxBest value$2.50
GPT-5.6-sol$6.25
Claude Opus 5$6.25
Claude Fable 5$12.50
$0$6.25$12.50

One-time cache creation on Qwen is 60% below sol and 80% below Fable 5.

Full pricing detail

USD per 1M tokens
Model Input Cached input Cache write Output Long-ctx input Long-ctx output
Qwen 3.8 MaxBest value $2.00 $0.25 $2.50 $6.00 Included — 1M context at base price
Claude Fable 5 Anthropic $10.00 $1.00 $12.50 $50.00
Claude Mythos 5 Anthropic $10.00 $1.00 $12.50 $50.00
Claude Opus 5 Anthropic $5.00 $0.50 $6.25 $25.00
GPT-5.6-sol OpenAI $5.00 $0.50 $6.25 $30.00 $10.00 $45.00