Flagship — best answers, highest costOpen Weights

Qwen3.8 2.4T A95B (batch) Pricing

Qwen · qwen-qwen3-8-2-4t-a95b-batch

ShareX Facebook LinkedIn

Input / 1M tokens

$2.00

Output / 1M tokens

$6.00

Context window

1010k tokens

≈ 1,347 pages of text

Max output

909k tokens

What it costs in practice

Typical request (1,200 in + 400 out tokens)$0.0048

1,000 requests / month$4.80/mo 10,000 requests / month$48.00/mo 100,000 requests / month$480.00/mo

Estimate your own workloadOpens the calculator with Qwen3.8 2.4T A95B (batch) preloaded — adjust volume and token counts there.

Price history

List price per 1M tokens since we started tracking in Aug 2026. Each step marks a price change — hover a point for the exact date and rate.

What it excels at

Supports 1M-token context and 900k-token outputs, enabling large-scale batch processing of long documents or multi-step reasoning.

The business tradeoff

Open-weights deployment introduces hosting variance and infrastructure costs; output pricing is 2.5x input with limited ecosystem tooling compared to managed APIs.

Cheaper from Qwen

Qwen3.7 PlusQwen$0.0009/ typical request Qwen3.8 27BQwen$0.0017/ typical request Qwen3.8 FlashQwen$0.0004/ typical request

Same tier from other providers

Claude Fable 5Anthropic$0.032/ typical request Grok 4.3xAI$0.0025/ typical request DeepSeek R1DeepSeek$0.0015/ typical request

Head-to-head

Qwen3.8 2.4T A95B (batch) vs Claude Fable 5 →Qwen3.8 2.4T A95B (batch) vs Grok 4.3 →Qwen3.8 2.4T A95B (batch) vs DeepSeek R1 →

Vendor list rates, as of Sep 1, 2026 · source: openrouter · per-request examples assume 1,200 input + 400 output tokens.